DOCUMENT RESUME

ED 089 559 HE 004 982

AUTHOR Cope, Robert G., Ed.TITLE Tomorrow's Imperatives Today. Proceedings of the. 13th

Annual Forum, Vancouver, British Columbia, TheAssociation for Institutional Research.

INSTITUTION Association for Institutional Research.PUB DATE 6 73NOTE 202p.; Proceedings of the Annual Forum of the

Association for Institutional Research (13th,Vancouver, British Columbia, May 1973)

AVAILABLE FROM The Association for Institutional Research, 201School of Education, Florida State University,Tallahassee, Florida 32306 ($8.00)

EDRS PRICEDESCRIPTORS

MF-$0.75 HC-$10.20 PLUS POSTAGE*Administrative Problems; Conference Reports;*Educational Administration; Educational Research;Faculty; *Higher Education; *Information Processing;Institutional Research; *Management; ResourceAllocations; Students

ABSTRACTPrepared from addresses and papers selected from

those presented at the' 13th Annual Association for InstitutionalResearch Forum in May 1973, this collection deals both with the broadnational issues facing postsecondary educationl'and with matters thatare directly related to campus functions, such as computer simulationmodels. The tone of these presentations, reflecting a growinguneasiness, is not optimistic. The contributors on topics dealingwith management, data systems, resources, faculty and students, saythere is at least the need for improvement if not reform. They haveanalyzed the problem from their perspectives and suggest thedirection change must take. The papers are divided into sixcategories: Imperatives in postsecondary Education; ManagementImperatives; Information Imperatives; Resource Imperatives; FacultyImperatives; and Student Imperatives. (Authors/EG)

w

1`- , Ct2P1

VA f : Eh,

, .

:

v

'.tQsii""F * 1E,T

U S DEPARTMENT Of psEALTH

FOuCAT,ON AESF ARES4h.1014/..s...NSTTuT£ OF

EOLICATION;- PEE,.

,AN IAA .

E

,..-E s.f

,F. C:c

Proceedings of the 13th Annual ForumVancouver, British Columbia

The Association for Institutional Research

a

TOMORROW'SIMPERATIVES

TODAY

Edited by Robert G. CopeUniversity of Washington

Proceedings of the 13th Annual ForumVancouver, British Columbia

The Association for Institutional Research

Annual Forums on Institutional Research

1963 The Role of Institutional Research in Planning1964 A Conceptual Framework for Institutional Research1965 Design and Methodology in Institutional Research1966 Research on Academic Input1967 The Instructional Process and Institutional Research1968 Institutional Research and Academic Outcomes1969 The Challenge and Response of Institutional Research1970 Institutional Research Lind Communication in Higher Education1971 Institutional Research and Institutional Policy Formulation1972 Reformation and Reallocation in Higher Education1973 Tomorrow's Imperatives Today

Copyright 4 1973 by the Association for Institutional Research

Preface and Acknowledgements

Tomorrow's Imperatives Today, prepared from address and papers selected from those presented at the 13th Annual Forumin May 1973, deals both with the broad national issues facing postsecondary education and with masters, like computersimulation models, that are directly related to campus functions. The tone of these presentations, reflecting the growinguneasiness, is not optimistic. The prevailing tone is one of urgency.

The contributors, on topics dealing with management, data systems, resources, faculty and students, are saying there is atleast tred for improvement, if not reform. They have analyzed the problem from their perspectives and suggest the directionchange must take.

The preparation of this volume was aided greatly by the advice of Clifford Stewart, by the careful preparation of contributors'manuscripts. and by the service of a team of conscientious reviewers: Paul Bradley, Jr., Mary fo Clark, Robert Grose, BertrandHansen, Robert Pardon, Marvin Peterson and Gary Stock. Our hope is that this publication will advance the art ofinstitutional analysis, at least a little.

University of Washington Robert G. Cope, EditorOctober 1973

CONTENTS

PREFACE AND ACKNOWLEDGEMENTS

IMPERATIVES IN POSTSECONDARY EDUCATION

THE NEW VIGILANCE: 1

A.W.R. Carrothers. University of Calgary

GOALS AND THE FINANCING OF HIGHER EDUCATION 7

Howard R. Bowen, Claremont University Center

THE CHOICE OF A FUTURE 15

Joseph T. Sutton, University of Alabama

HIGHER EDUCATION UNDER STEADY-STATE CONDITIONS 18

Allan NI, Cartier. Carnegie Commission on Higher Education

THE HAZARDS OF PLANNING PREDICTING PUBLIC POLICY 23

Robert J. Kibbee, City University of New York

MANAGEMENT IMPERATIVES

PROGRAM EVALUATION: APPROACHES AND PROCEDURES 27

Cameron Pincher, University of Georgia

BET rER PROGRAM EVALUATION: HOW CAN WE MEET THE CONCERNSOF TODAY WHICH WILL BECOME THE IMPERATIVES OF TOMORROW? 34

Oscar T. Lenning, American College Testing Program

A DETERMINISTIC MODEL FOR INVESTIGATING THE FEASIBILITY OFYEAR ROUND OPERATION 39

C.J. Christie and D.J. Hipgrave, Woods. Gordon & Company

ACCOUNTABILITY: SUBJECTIVE IMPERATIVE 51

Bernard Trotter and Mario Creel, Queen's University

iNsTramo NAL ACCOUNTABILITY CONCEPT, DOGMA AND REALITY 57

Sibley. University of Manitoba

THE INSTITUTIONAL RESEARCH IMPERATIVE AT EMPIRE STATE COLLEGE 64

A. Paul Bradley. Jr. and Ernest G. Palola. Empire State College

INSTITUTIONAL RESEARCH AND THE EXTERNAL ENVIRONMENT: AN APPROACHTO CONTEXTUAL ASSESSMENT 69

Paul R. Lyons. Frostburg State College

INFORMATION IMPERATIVES

PLANNING MODELS IN HIGHER EDUCATION: A COMPARISON OFCAMPUS AND RRPM 74

K.M. Hussain, New Mexico State UniversityThomas R. Mason. University of Colorado

THE IMPLEMENTATION OF CAMPUS/COLORADO AT THE UNIVERSITY OF COLORADO... 86

Madelyn D. Alexander. University of Colorado

SIMULATING THE UNIVERSITY AN EVALUATION OF EXPERIENCE WITH THECAMPUS, HELP/PLANTRAN AND SEARCH SYSTEMS 92

Jerome F. Wartgow, Governors State University

RESOURCE PREDICTION MODELS: A DO-IT-YOURSELF APPROACH FORINSTITUTIONAL RESEARCHERS 96

Webster C. Cash, Fisk University

PLANNING-MANAGEMENT SYSTEMS IN HIGHER EDUCATION: WILL THEYREALLY MAKE A SIGNIFICANT CONTRIBUTION To IMPROVE INTERNALDECISIONS? 101

Peggy Heim, Bucknell University

THE COMPARABILITY QUES*110N: POTENTIAL USES AND MISUSES OF DATA 103

Adrian H. Harris, University of California. Los Angeles

AN INFORMATION BASE FOR BLACK HIGHER EDUCATION 108

lames Welch. Institute For Services To Education

RESOURCE IMPERATIVES

FINANCING UNDER CONDITIONS OF DECLINING FLEXIBILITY 111

James R. Montgomery. John Janney, Gerald W. McLaughlin, andDavid Sieg, Virginia Polytechnic Institute

UNIVERSITY IMPACT ON LOCAL ECONOMY 118

Bernard S. Sheehan and 13arbara Serekiak. University of Calgary

THE USES OF COST DATA IN INSTRUCTIONAL PROGRAM EVALUATION 000

Robert D. Lamson, Boeing Computer Services

FACULTY IMPERATIVES

A NEW LOOK AT TENURE 128

Thomas M. Freeman and Joseph G. Rossmeier, Michigan State University

THE TEACHING EFFECTIVENESS OF AN ALTERNATIVE TEACHING FACILITY 137

Peter Horowitz and David Otto, University of Alberta

RESEARCH REPUTATION AND TEACHING QUALITY IN UNIVERSITY DEPARTMENTS 144

Judy Richardson, Terry Eade and John McMillin, University of Washington

SEXUAL BIAS ANALYSIS IN HIGHER EDUCATION: AN APPRAISAL OFMETHODOLOGY USEFUL TO INSTITUTIONAL RESEARCHERS 150

Michael J. LaBay and Randolph N. Foster, Youngtown State University

BIAS IN FACULTY REPORTS OF TIME AND EFFORT EXPENDITURE 156

Patrick H. Sullivan, Florida State University

DETERMINING FACULTY ATI1LUDES TOWARD COLLECTIVE BARGAINING USINGPATH ANALYSIS 162

Malcolm C. McInnis. Jr., University of TennesseeF. Craig Johnson, Florida State University

STUDENT IMPERATIVES

AN ANALYSIS OF THE EFFECTS OF TUITION AND FINANCIAL AID POLICIES 170

Robert Winter, University of Illinois

THE EDUCATIONAL ATTITUDES AND ORIENTATIONS OF UNDERGRADUATES: ANEW MEANS OF ASSESSMENT 178

Barry R, Morstain. University of Delaware

COLLEGE DROPOUTS: THEORIES AND RESEARCH FINDINGS 185

William P. Fenstemacher. Minnesota State College System

THE NEW VIGILANCE

I accepted the honour of your invitation to give thiskeynote address, not out of any conviction that I have any-thing new to say to you, but because I detected in theinvitation a recognition of the contribution being maderegionally, nationally, and internationally by the Office ofInstitutional Research at the university of which I happento be executive head. That office was created in my fresh-man year as President not because I had any great vision ofbringing the science of management to solving the myriadproblems of today's universities, but because I waspersuaded by our Director's advocacy that such an officecould play a major role in bringing peace, order and goodgovernment to an institution the condition of some ofwhose affairs seemed at that time to range from the ridi-culous to the chaotic. Since then 1 have been persuaded onmatters relating to the development of that office, again notbecause I understood the case at any technical level butbecause I was persuaded by the record of the office thatinstitutional research was a sound investment in ration-alizing structures, systems, and decision-making processes.A modern university in a modern world needs amechanism that will provide relevant and accurate infor-mation to and about management. What I would like to talkabout tonight is just that: the nature of the modern worldwhich plays host to its universities; its significance touniversities; the responses which an alert institution mustbe prepared to consider; and what that means for the rolewhich offices of institutional research will be called uponto play.

A keynote address, as I understand it, is not sup-posed to say anything new to the conference participants.It should as best anticipate in a prosaic way the new ideaswhich the participants will reveal with professional clarityas the conference progresses. I come from a profession thatspends much of its time in the past in the hope of antici-pating the present. Your profession lives largely in thefuture, using what you can of the past rnd the present todetermine what may be, in aid of those who must judgewhat should be. I therefore stand before you merely as auser's committee of one, I realize that this is an inter-national conference. I hope you will forgive me if I talkabout the country with which I ought to be familiar, Theproblems may not be that different, but I do not presume tolecture others on their affairs.

Carrothers, University of Calgary

University Dependence on Tax DollarsOf the 64 member institutions in the Association of

Universities and Colleges of Canada I doubt that there arehalf-adozen which have endowments of sufficient sub-stance that they can consider private dollars as providingany significant discretion in planning and operating theiraffairs. The rest overwhelmingly are dependent on taxdollars for their capital and operating budgets; even theproportion of income from student fees is going down as, inmost jurisdictions, these stay frozen while costs rise. In thiscountry, education is a provincial constitutional respon-sibility, and nearly all universities are creatures of provin-cial statute. Since 1967 the federal government hasreimbursed provincial governments 50% of operating costsincurred in all post-secondary institutions. This very sub-stantial federal commitment is based on the view thatuniversities are much more than educational deliverystations: that they have an important role in research, inother matters of national interest, and in the performance ofnational responsibilities in international and world affairs.In short, collectively, they constitute a very considerablenational asset.

The federal government had been involved in directfinancing of universities on a per capita basis of provincialpopulation for some 15 years prior to the 1967 cost sharingarrangements, The introduction of the new open-endedscheme coincided with a number of changes that reinforcedone another to push universities into almost a messianicrole. You are familiar with them: the explosion of basicscientific knowledge in the post-war period symbolized bysputnik; the rise of a new technology, symbolized andrepresented by the computer; the emergence of the socialattitude that education is the key to our salvation; the con-sequent inflation of individual expectations; the rise in thepercentage of young men and women of university ageseeking access to universities known to the boffins asthe participation rate."

As North America turned into the present decade,dragging behind it an economic recession from the tag endof the 1960's. a backlash against universities set in. First, asI perceive it, came general alarm over the escalation ineducational costs, encouraged by prognosticators who pro-jected educational costs as rising above gross nationalincome before the turn of the century, coupled with

1

NEW VIGILANCE

criticism from both business and labour that the world oflearning was inappropriately remote from the world ofwork. Then came disillusionment from university graduateswho could not find employment to their particular tastes.Then came an unprcdicted levelling-off in enrollmentI suppose what might be called a stabilization of the partici-pation rale. Then came a crisis in university managementbecause institutions that were planning for growth andwhose income related directly to enrollment could notbalance their budgets. University dependence on taxdollars was no longer an hypothesis or an assumption orsomething one lived with, like earthquakes, birthdays, andother Acts of God; it was a brutal reality. Governmentswere not inclined to be moved by this new plight of theuniversities. Although informed members of the public.when they are reminded of it, are in my experiencebasically sympathetic to the unique role which universitiesperform in society, when it comes to the allocation of taxdollars they are very suspicious that there is fat in ourbudgets, and their suspicions are reinforced by apprehen-sions or misapprehensions respecting such matters astenure; sabbatical leave; teaching loads; the quality ofresearch and the motivation for research: the comparativeneglect of freshmen and sophomores: and the fact that theone thing that universities clain, they cannot account for ishow a professor actually spends his time. I think also thatthere is recognition, more widespread than faculty mayrealize, that the freedom, of the academic community protects other freedoms. and that a threat to the formerdeserves a response in the cause of all freedoms. Yet ifthose who influence public opinion believe that univer-sities are in need of reform and are unconvinced thatuniversities have reformed, they will not put their weightbehind the universities' plea for help over a mere financialcrisis.

At the present time universities are struggling to findtheir place within a new range of choices among post.secondary institutions in an atmosphere of popular interestin the institutionalization of "life-long learning," a concernwhich could itself have a distorting impact on educationalinstitutions and selection of goals.

Increase in Government Interest in University AffairsAlong with the increase in university dependence on

tax dollars has come an increase in the level of curiosity bygovernments about what universities do in fact. Govern-ments are interested in academic programs. in the rational-ization of graduate studies, in research policies and in howwe manage our affairs. A new set of initials has emergedH.Q.M. Highly Qualified Manpower whichsymbolizes government expectations of universities as theyrelate to economic interests and particularly to the market-place for university graduates. Universities have alwaysbeen repositories of highly qualified manpower. andgovernments have frequently called upon universitypersonnel to advise on a wide range of economic and socialissues and to represent governments in international

2

affairs, Now governments want to know what universitiesare doing on their own initiative and how their productionof graduates relates to social need. That is the "new rele-vance.- In the last t2 years virtually every jurisdiction inCanada has had a public enquiry and report on education,from the Parent Commission in Quebec to the Wright-DavisReport in Ontario. People now speak freely of universitiesas public utilities, and of education as a right. Yet there isconflict with that quality of autonomy which is essential tothe performance of the universities' unique role in thequest for truth and between the individual's interest in aright to an education and the slate's interest in highly qual-ified manpower planning.

Most important of all. other social needs competewith education for the tax dollar, and heightened govern-ment interest in universities relates profoundly to thenever-ending redetermination of public priorities. Educa-tion is a kind of experiment that failed to come off, andgovernments are not disposed to repeat it; they are lookingelsewhere for success.

The Challenge to AssumptionsThe environment of institutional dependence on

substandial tax dollars and heightened government interestin what universities are doing, and therefore in what theyshould be doing. presents a challenge to virtually everyconventional assumption about universities and their rela-tions with governments.

Orthodoxy asserts that certain matters are the prero-gative of governments, including the prerogative of creatingpublic universities. By the same token, once a university iscreated, orthodoxy asserts that it takes on a quality tailoredto its particular assignment of the quest for truth, whereverit may be found and in whatever direction it may takesociety, and the university has assigned to it an estate.that quality of independence, known as autonomy, com-mensurate to that task. Governments and universities, eachin their own way, claim to serve the public interest. Never-theless, they never really have lived in worlds of their own,and they are now, perhaps more than ever before, calledupon to share the same path. May I be more specific.

Basically the allocation of public resources to univer-sities is a government responsibility. Should universitiesbe free to seek access to private resources? Has the privatesector an interest in influencing the profile of publicuniversities? Have governments a legitimate interest in theimpact of such resources on the profile and capacities ofpublic institutions? Should universities be free to assessstudent fees. or have governments a legitimate interest ineducation costs from the student viewpoint, to the extentthat governments not only should control fees but alsomight consider the degree to which moneys conventionallymade available fairly directly to institutions should bemade available through the hands of the student? Shouldstudents, as they vote with their feet, carry tax dollars intheir pockets?

Who should go to university? Traditionally. univer-

sities have determined admission standards within thepolicy of the open door. With the growing abolition of highschool leaving examinations as a result of governmentpolicy, we may be moving to a completely open-doorsystem, in which the new expectation may be that everyoneis entitled to be an unqualified success. Will the govern-ment pay the cost? Should the interest in a new controlmechanism over quantity and quality, which is inevitable,be perceived as unilateral or as a shared responsibility?

Heretofore governments have shown littk direct in-terest in how public monies, once allocated, are in factused. Universities determined the internal allocation togeneral undergraduate studies. to graduate education andresearch, and to vocational or professional training. I amsure as you heard those words you recognized in these mat-ters a public interest of which governments may in-creasingly view themselves as stewards.

Not long ago universities were the sole occupants ofthe post-secondary educational field. They could strive tobe all things to all men, for they were all that men had. Infact the great strength of public universities in the westernprovinces. like the land grant colleges in the United States,lay in their close identification with the needs, desires andambitions of large numbers of citizens in the post-highschool level of education. Now universities are challengedto identify their particular role among a number of kinds ofpost-secondary institutions. We can no longer be all thingsto all men; if we try we will fail, and will be blamed for it.Should we now be planning a retreat to the ivory tower?Universities must recognize the issue of identity as theirchallenge; but again, we do not operate in isolation, if forno other reason than that what we can do may depend onwhat public resources are made available to us.

I am no longer very confident, when I use the termsshort-. medium- and long-range planning. because I lackconfidence in my judgment of their time spans. Yesterday'slong range is often today's medium range and tomorrow'sshort range. But whatever long-range planning is, whatevermay be the interest of the university in it, there is today anobvious government interest in it as well. The great un-certainty at the moment is how much planning is therelikely to be or can there be and how is one to select the ob-jectives to which the planning is to be directed. The casefor shared responsibility is increasingly obvious.

Finally, we must reexamine the barricades of institu-tional autonomy and academic Freedom, Are we any longerfree can we any longer afford to be free to determinehow to teach, how to carry out research, and how toorganize ourselves to these tasks? Is there now a social in-terest in these matters as there is in what programs are tobe offered. in the participation rate al both the undergrad-uate and graduate levels, and in the nature and extent ofresearch? Have we not lost something of the legitimacy of

Carrothers

decision-making in these matters that we used to enjoy? Ifwe are right, as I believe we are, in continuing to insist thatautonomy is an essential concomitant of our unique socialrole in the quest for truth, a role which gives universities alife force of their own, is our stronger claim to autonomyand freedom in the areas of what, how, and whom to teachnot that by nature they are ours absolutely but that we canbetter exercise judgment in respect of them'? if so, must wenot stand ready to account for our judgment? We are sensi-tive to the fact that in internal university affairs leaders areconstantly called upon to authenticate their leadership bythe quality of their judgment, and that nothing is acceptedas settled so long as the merest morsel of argument hasbeen left unsavoured.

The Byzantine Condition of Post-Secondary RelationshipsAn important component in the Canadian education

environment today, in addition to the challenge to conventional assumptions, is the pluralistic nature of interests.The British North America Act allocates education to theprovinces, at least in and for the provinces, whatever thatmay mean; and provinces create educational institutionscalled universities. These in turn engage in activitiesbeyond "education" in the constitutional sense, activitiesthat naturally, properly and necessarily attract the interestof the federal government. Provincially created institutionscongregate in a national association that is created, incor-porated and regulated under the aegis of a special statute ofthe parliament of Canada. Even two-year "community"colleges are formed into a national association. Facultyspeak at provincial, regional and national levels throughtheir voluntary associations. Provincial ministers of educa-tion form themselves into a council with a central ornational office. Research programs of universities, parti-cularly in the natural and health sciences, are heavilyinfluenced by national funding agencies and governmentdepartments, a matter that increasingly is attracting theinterest of provincial governments in respect of programs,financing. and "rationalization," Research of course isinterrelated with graduate studies and the production of ournew acquaintance "highly qualified manpower," clearlymatters of shared interest. Even learned societies, whichheretofore functioned as genteel and scholarly academicguilds, are in particular cases politicizing themselves andsetting themselves in judgment over certain forms ofmanagerial activity.

The Internal Impact of the New EnvironmentThe new challenge to old assumptions and the

pluralistic nature of the educational environment arehaving their impact on the internal affairs of universities.

To begin with, faculty today feel threatened as theyhave never felt since the depression of the 1930's, which

3

NEW VIGILANCE

means for practical purposes, such being the nature of thememory of man, vs they have never felt before. They havebeen exposed to the policy of shareJ responsibility andthey are disillusioned with it, for the results are not all thatdifferent from the bad old days of ''authoritarian" rule.They are affected personally by the institutional identitycrisis. It is a s'lort distance from being threatened in one'semployment to questioning the worth of the job. Facultyare mindful of the uncomfortable trade-offs at stake andtheir impact (4) individual economic and professional wellbeing.

Their response is a highly predictable surge in thedirection of collective bargaining. Whether collectivebargaining comes as a matter of volition or as a matter oflaw, a great many things must be sorted out. When themechanics are squared away, issues of mutual trust havebeen resolved, and the real bargainers identified, there stillremains the fact that effective collective bargainingdepends on the availability of data that both parties acceptas relevant, reliable and complete. I shall return to this in aminute.

At the level of institutional management the chal-lenges of the new environment are shaping behaviour.Undoubtedly hard times have a centripetal impact. So doesthe policy of planning. So does the policy of accountability,for no sane person will accept responsibility for a matterbeyond his control. So does the wilfullness of politics,because in an environment of unpredictabilityindividualism will be surrendered for the protection to besought in the herd.

I think it overstates the response to assert. as somewould, that there has been a failure of nerve within institu-tions. I think it may be fairer to say that they are drivingdefensively. There certainly has been a loss in the politicalbase of institutions. There also has been an escalation inthe rate of turnover in administrative personnel thatfollowed the adoption of term appointments in conjunctionwith the mechanisms of shared responsibility betweenfaculty and administrators in the mid-1960's. The questionmust be faced whether there is an adequacy of leadershiptrained to the problems of politics, collective bargaining,systems management and public accountability, combinedwith the competing demands of "shared responsibility"and conventional academic management and, hopefully,academic leadership.

Two matters are of greatest concern to universities inthe 1970.s. One fear is that they will be reduced to a subsis-tence level of performance because of a substantial reduc-tion in public priority given to this component of highereducation. The other is that universities will be politicizedas a consequence of a substantial rise in the level ofgovernmental interest in what they do in fact, not simplybecause of the demand for accountability of governments,which inevitably passes on to ;ax supported agencies and

4

institutions, but also because of a rising interest by govern.ments in the utilization of the expertise reposing in ouruniversities in the search for solutions to the increasinglycomplex issues facing governments today.

The Need for InteractionThe convoluted nature of our relationships is not

going to be resolved, and the modern challenges to conven-tional assumptions are not going to disappear, We need tolearn to manage our environment. or at least to manage our-selves in our environment. If universities have lost ameasure of public confidence and hence a measure of theconfidence of governments, so also is there rising appre-hension within universities that they are to be exposed toprejudicial government policies created without an oppor-tunity for them effectively to influence the determination ofpolicies and theft execution. That is the current "crisis ofconfidence." It is basically a crisis of communication. I

have on another occasion used the analogy of a roomful ofcuckoo clocks; when each clock thinks its time has come abird pops out, has its say, and pops back in, quite indif-ferent as to whether it has been heard and quite unin-terested in listening to others. Governments, universitiesand the public whose interests we both claim to servedesperately need to develop a philosophy of communi-cation, a commitment to interaction, and from that todevelop processes, or, to use current planning argot, an"effective interface." through which assumptions can bereviewed and progress made toward identifying and attain-ing long term objectives, by which we can influence ourfuture. We cannot live by hidden controls and crypticmessages. Governments and universities together mustresolve the uncertainties and minirnize the risks of presentunstable conditions. Universities must be prepared to dotheir homework, and governments must be prepared totreat them with candour. Neither governments nor univer-sities can afford to be taken by surprise.

Above all there is need for mutual recognition of theproblem of processes as it affects governments and univer-sities today. Where there is recognition of objectives, how-ever easy or uneasy those objectives may rest on thepartips, it is comparatively easy to deal with processes. Butwhere assumptions and objectives are subject to seriousquestioning, there is put into issue not only the rote ofuniversities, not only the role of the state, but the wholerole of education in society over whatever planning spanwe may think we have or may in fact be given to us. Theproblem of processes therefore is the problem of how theparties are to live together in their condition of unsettle-ment and how the parties are to manage the interrelation-ship of processes and the identification of educationalobjectives and their priorities through the reconciliation ofcmpeting interests. Throughout must be recognized therisks attendant upon planning not merely that planning

will be bad in the sense of less than competent, but that theclearer the planning the more obvious who will have andwho will have not; who will benefit from the plannedcondition and who will perceive his destiny in rebellionagainst it, rebellion, hopefully, within the political processitself. Throughout must be recognized the disparate natureand function of governments and universities in our kindof society, and the legitimate role of institutional autonomy.

The ad hoc committee on planning of the Asso-ciation of Universities and Colleges of Canada, of whichmany of you are members, was created toward the end ofthe 1960's with the intent of being concerned primarilywith physical planning. When the boom went out of uni-versity growth the committee, at a conference at LakeheadUniversity in 1970, which was inward inoking and philo-sophical in nature, turned to the interrelationship ofphysical and academic planning. A further conference washeld at the University of Calgary in 1972 at a more technicallevel, addressed particularly to the development of manage-ment information systems. In November, 1972, the com-mittee met with its advisory committee, and shortly there-after it brought to the Board of Directors of the Associationa three-phase plan for grappling with the problem ofdeveloping an effective interface with governments. TheBoard adopted the plan enthusiastically. Phase one, whichcalled for executive action, was completed in short ordr.:Phase two took place in mid-March of this year. It consistedof a workshop of some two dozen persons, drawn fromfederal and provincial governments, universities, businessand labour, to consider in a comparatively unstructuredway these issues relating to the development of processes.Many of the points I have sought to make today came outclearly in that very intense and stimulating day and a half. Iknow I gained from it.

The workshop resolved that as a next step thecommittee should "assess the processes for planning longterm development for universities in Canada with specialattention to interactions of institutional, provincial, regionaland national authorities." That step should be completedshortly, and as President of the Association I have askedthe committee chairman to couple his report with proposalsfor the next phase in the exercise. I tell you this becausethink it is worth noting than an ad hoc committeeoriginally created to concern itself with physical planningexpanded its mission to include the union of academic,physical and fiscal planning, and was instrumental ininciting the Association of Universities and Colleges ofCanada to provide leadership in grappling with the newand critically important problem of processes.

The Role of Institutional ResearchI have been talking about the processes of planning

within universities, between universities, within govern-ments, between governments and between governments

A.W.R. Ca rto th ers

and universities. That is emerging as the educationalchallenge of the 1970's. If it is not approached at the highestlevel of sophistication the game will be rough and publicopinion may be a poor referee. Some may have no stomachfor the game, which is understandable. Others may findtheir durability tested. Participants who are highlydependent on Job satisfaction for a sense of personal well-being may be hard pressed to find alternative sources. I seeinstitutional research in the 1970's as playing a major roleof reinforcement to those who are charged with developingan effective planning interface between governments anduniversities. You are the staff, the boffins, the experts. Youneed not be reminded that you are not the policy makers;but you are the constants in the process of policy making,and you can be of major use in reminding administrators ofwhat they need in supplying policy. We must be masters inour own house; we must be masters of the facts about our-selves, for only then will we be able to demonstrate ourcompetence and thus to legitimize our claim to autonomy.At the political level you have no visible role, but again youcan and must back up administrators with accurate andrelevant information.

Until now institutional research has been meeting aninstitutional demand fcr costing services respecting theacquisition and allocation of capital and operatingresources. It has served a need for systems development notonly at the institutional level but at provincial and federallevels as well. It has met myraid ad hoc requests. and inthe process it has taught administrators a thing or two aboutasking the questions they really want answering and askingthem in art answerable form. I personally have the greatestrespect for your collective perspicacity and accomplish-ments.

think you now face a number of specific demandsas you practice the applied science of systems engineering,

PPBS (Program Planning Budgetting Systems) islikely to be with us for a very long time. It requiresparticular adapation to university management, but govern-ments have adopted and adapted it and we must continueto do the same. It will, as I see it, involve the process ofprediction over a wide range of facts; the development ofmore and more sophisticated allocation models foroperating resources; and the continuing refinement of thetechniques of cost benefit analysis.

You doubtless will be called upon to provide datarelating to the inter-institutional rationalization of researchand graduate studies. and once again you may have to teachyour interrogators how to ask the right questions.

The move to collective bargaining could well shift aspecial burden to institutional research. The negotiation ofterms and conditions of employment is likely to raise issuesquite distinct from those to which institutional researchersare accustomed. Perhaps the most important matter iswhom are you there to serve, You understand now that

5

NEW VIGILANCE

your duty is to apply your professional competence andstandards to the institution, and you need not distinguishbetween the components of the institution. The answer to aquestion respecting a cost study does not depend on theidentity of the questioner. If you can preserve that virtue inresponding to demands for information to be laid on thebargaining table, you will be performing the best possibleservice to the institution as a whole. If facts are to resolveissues at the bargaining table they must be above reproach.On this I shall say no more for fear of saying too little.

Institutional research has always been an importantcontributor to academic planning. Today universities faceparticular issues in academic planning in the areas ofcontinuing education and the concept of the open univer-sity. Here the university confronts a broad spectrum ofsocial needs, from the conventional academic program, toprofessional retraining, to learning for leisure. We facemajor policy decisions. The burden will be made lighter bythe availability of information respecting the educationalmarketplace,

There is a growing demand for managementdevelopment programs for academic administrators. It hasbeen my observation and experience that management con-sultants, in bringing their talents to bear on educationalmanagement problems, are extremely useful, for a professionally trained outsider can perceive deficiencies anddesign solutions which the insider cannot grasp on hisown. But their models, which tend to be derived in the firstinstance, and often directly, from private enterprise, a verynatural prototype for management science, will bend onlyso far to the needs of universities. The institutions mustreach out to bridge the gap between the capacities ofprofessional consultants and the particular needs of educa-tional institutions, Those in institutional research may becalled up to do just that.

Your particular forte is the development ofmanagement information systems. Information is the first

step to planning. Planning implies control, and controlimplies a pecking order. You will be in the eye of thishurricane of systems development: and you will be calledupon to share the product of your skills with other insti-tutions, and to develop a coordinating apparatus.

Earlier in this address 1 referred to a number ofconventional assumptions that are being challenged, I alsosuggested, more than once, that universities should take theinitiative in stating the issues and proposals for theirresolution, The universities should accept that onus andstate their policies on such matters as size, participationrate, programs, financing and similar matters which add upto a statement of fairly concrete objectives. That at leastwould give a basis for dialogue with governments, and inthe performance of the exercise it is conceivable that someproblems might be identified and resolved and a number ofquestions might disappear. Then we will know that theresidue contains the issues that need a lot of hard work.

I have been talking about the challenge9 and thecrises of the decade of the 1970's. A hundred years fromnow what will be the centre of the university will not bethe problems of administrators or the role of institutionalresearch, It will be the liberal arts not rrmlicine, not law,not engineering, but the study of the nature of man.Furthermore, education is an instrument of social progress,and that will carry the day. The unique role of the univer-sities will still be the quest for truth. That quest calls forrationality, which in turn establishes the role of rationalityin management. That is the particular role of institutionalresearch: that and the understanding of the limitation ofrationality, for you must not let us quantify the unquanti-fiable. Institutional research is the new buffer betweenuniversities and the shifting forces of politics. "The price ofliberty is eternal vigilance;" and a university must be "aplace of liberty:' We look to you, the professionalrationalists, to bring a new vigilance to the challenges ofour time.

GOALS AND THE FINANCING OF HIGHER EDUCATION

For several years, I have been brooding over thequestion of how higher education ought to be financed. Ihave been struck by disparity between American practiceand that in most parts of the world. Most nations admit totheir universities a smaller percentage of their populationthan we do, but once having admitted them they offerhigher education without tuitions and often with heavysubsidies for living expenses. In the United States, on theother hand, we are constantly raising tuitions and areshifting more and more of our student aid from giants tolong-term loans.

I have also been struck by what seems to be a radicalchange in American attitudes about higher education. Forat least a century higher education has been regarded as aform of personal opportunity for the sons and daughters oflow-income families. It has been thought therefore that itought to be made available on the most generous terms.This, I have thought, was the idea underlying the MorrillAct, the establishment of the public urban colleges, thecommunity college movement, and the GI Bill. I myselfattended a state college (not far from Vancouver) whichhad open admissions, no tuition, and living costs so lowthat virtually anyone could get through. Why, I have asked,at this particular stage in our history when we still have thetask of bringing millions of young people into the mainstream of American life many of them from ethnicminorities are we suddently shifting to high tuitions andlong-term loans? Have we been misguided over the yearsand are we just learning about our mistakes? Or are wenow committing a colossal blunder?

The puzzle is perhaps especially acute for mebecause many of the members of my own profession, econ-omics, have been among the chief advocates of hightuitions and student aid in the form of loans. Thier view-point derives in past from the economists' traditional loveaffair with the market as an allocator of resources and theirsuspicion that the finance of higher education through thetax system represents an adverse redistribution of income.

As I have pondered these questions over the pastfive or six years, my ideas have evolved considerably. Ihave come to realize that the issues in the financing ofhigher education are primarily matters of value rather thanof technique. The higher education system serves a multi-

Howard B. Bowen, Claremont University Center

plicity of goals. Some of these goals are mutuallyincompatible az:d .some are assigned different priorities bydifferent persons. Each of these goals when viewed inisolation calls for a somewhat different kind of financing.When these goals are viewed collectively, they call for amixed system of finance not unlike that which currentlyprevails. The present system can be viewed as the productof the complex cross-currents of American politics. Thesystem is not tidy; it is based on no single ideology; it is fullof compromises; it is hard to understand; it fully pleases noone; it is likely to change through gradual evolution, notthrough radical departures. The basic policy question is: Inwhat general direction should change be tending?

In my remarks tips morning, I shall try to identifysome of the goals that are sought through higher educationand to explore their implications for financing.

Economic Growth and National Military PowerA major goal for higher education is nation econo-

mic growth. Many studies, as well as common sense. haveindicated that learning in a wide variety of fields addsto the productivity of labor and that the kinds of basic andapplied research conducted in universities also enhance thenational product. Moreover, the connection between learn-ing and national military strength has been widely recog-nized especially since World War H. Indeed, the revival ofinterest in higher education in the 1950's was sparked bythe launching of Sputnik; the major breakthrough in federaleducational legislation of that period was called theNational Defense Education Act; and one of the majorcomplaints of students during the height of the Viet NamWar was that higher education had become a tool of theMilitary Industrial Complex.

The goal of national economic growth calls forfederal financing of higher education partly in the form ofgeneral support and partly in the form of categorical grants,loans and contracts intended to encourage particular kindsof training and research. Categorical support results ineve: 7hanging emphasis in federal programs 83 old needsare fu,iilled and new needs discovered. it tends to putuniversities in the position of producing for a "market" inresponse to rapidly changing demands of the federalgovernment, and removes some of the decision-making

7

GOALS AND FINANCING

from the campus to Washington.

Supply of Professional and Other Skilled PersonsAnother goal is simply to provide an adequate

supply of professional and other skilled persons to servethe population. Society requires a reliable supply ofphysicians, dentists. nurses, teachers, lawyers, clergymen,accountants, engineers. soldiers, etc., just to take care of itselemental needs. Even in the founding of Harvard, thesupply of professional workers, specifically clergymen, wasa dominant consideration, and many of the state collegeswere originally founded as normal schools to provide asupply of teachers.

Historically, because professional people werescare and because some of the professions offered lowrenumeration, this goal called for a system of finance thatwould encourage students to enter the professions: low tui-tions and student aid in the form of grants. Today, exceptin the health fields, the scarcities have apparently beenlargely overcome, and in the health fields incentives arestrong because renumeration is high. So the need of specialfinancial arrangements to encourage the supply ofprofessionals is less strongly felt than it was in earliergenerations and this goal probably calls for no specialfinancial arrangements.

Citizenship and Civic LeadershipAn American article of faith is that widespread

education will produce an intelligent, informed, andresponsible electorate. The drive toward universal second-ary education nearly a century ago was promoted withcitizenship as a major goal. Today the claim is made thatnear universal education for two years beyond high schoolcan be justified on the basis of its contribution tocitizenship.

This position suggests that education in the first twopost-secondary years should be tuition-free or nearly so. Italso suggests that after the first two years the privatebenefit from higher education exceeds the social benefitand that substantial tuitions, or even full-cost tuitions,should rule for the upper-division, graduate, and profes-sional years.

This position also rests on the argument that oncestudents have experienced two successful years of college.have discovered their interests and abilities, and haveappraised their opportunities, they should then be able tojudge the private benefits of further educatio- and shouldbe willing to proceed under a system of nigh tuitionsbacked up by long-term loans.

This general view represents a neat compromisebetween those who would make higher education at alllevels essentially tuition-free and those who would chargefull-cost (or nearly full-cost) tuition at all levels.

Some uneasiness about the citizenship theory e uses

8

from the fact that many of those who favor wide extensionof higher education in the first two post-secondary yearsare thinking of vocational rather than general education.

The citizenship proposal can also be questioned onthe ground that citizenship may be advanced as much by ahighly-educated civic leadership as by mass higher educa-tion at grades 13 and 14. The interests of the body politicmight call not only for education on the part of the manybut also deep education on the part of the few (but not toofew). Society in seeking good citizenship may be justifiedin subsidizing education in the later as well as earliercollege years.

Solution of Social ProblemsA persistent goal of higher education is to provide

the knowledge and the personnel to cope with socialproblems. Today our society has an especially wide array offelt social problems most of which are ntaional in scopeand import. We wish to conquer poverty, achieve racialjustice, renew our cities, restore order, improve healtheducation, renew the environment, develop the arts, keepthe peace, restrain world population growth, and aiddeveloping nations. These tasks will require great bodies ofnew knowledge and great cadres of dedicated and profcs-sionally competent persons. They will stretch our resourcesin educated, sensitive, insightful people. Education andresearch are still our main hope for coping with theseproblems.

Solutions call for citizenship education of the manyand liberal and professional education in depth of theindespensable technical and political leadership. Solutionsalso require research and scholarship in the natural

.ces, the humanities, and the social studies.An adequate attack on our national social pro-

blems would seem to call for broad federal support andencouragement of all aspects of higher education, notmerely the first two years, not merely research in thenatural sciences, and not merely mission-oriented programsof education and research. But in addition to broad generalsupport, it would also require categorical aids to fosterparticular kinds of research and 'mining.

Responsiveness to Social NeedsThe allegation is often made that colleges and univer-

sities depend upon public support but are not responsive oraccountable to any constituency. It is said that they areinsensitive to the needs of the government, to the interestsof their students, to the wishes of the various publics theyserve, and to the broad social interest. II is argued thattheir governance and finances should be changed to makethem more responsive.

No one can reasonably deny that colleges anduniversities exist to serve their society and that they have adeep obligation to work toward advancement of the

genuine interests of that society. The question is: Now canthis social responsiveness be best assured?

One way that is often advocated is to finance institu-tions through the price system. Instruction would be finan-ced through tuitions received from students; auxiliaryenterprises would be financed through fees to cover costsincluding capital costs; public service activities would befinanced through payments from individuals, corporations.and public agencies receiving services; research andscholarship would be financed through grants or contractsfrom individuals, corporations, and public agencies. In thisway, higher education would concentrate on those specificactivities which someone on the outside would deem worthpaying for. The distinction between proprietary and not-for.profit institutions and between public and private institutions would then largely disappear.

In recent decades, the system of higher educationhas moved perceptibly toward the market model. Contractresearch has become commonplace, tuitions and other feeshave risen and services of auxiliary enterprises areincreasingly priced at fullcost.

Another way of increasing responsiveness has beento alter the governance of institutions by subjecting them tosite visits, inspections, audits, and reports, by placing themunder the supervision of public coordinating bodies, andby direct legislation affecting programs, tuition, salaries,teaching loads, tenure, etc.

A question may be raised as to whether the"customers" on the outside are better able than the facul-ties and administrators on the inside to determine what isin the social interest. Concern for the long-run interests ofstudents is surely a responsibility of the college and univer-sity, but it does not necessarily follow that better educa-tional decisions will be niade if students control a largepart of the income of colleges and universities than ifeducational funds derive from unrestricted appropriationsand gifts. Similarly, research and scholarship shouldpromote the tongrun interests of society, but it does notfollow that better decisions about the development ofscience, the advancement of knowledge, and the enrich-ment of the culture will be made by outside individuals,public agencies, and corporations than will be made byfaculties. Even in the case of public services, it is not self-evident that outside groups are bound to be right about thekinds of public services that are compatible with the totalprogram and mission of colleges or universities. But to theextent that greater responsiveness to outside decision-making is sought, increasing finance through tuitions, fees,contracts. categoric& grants, etc., is one way of achievingthe goal.

Efficiency

In recent years, the opinion has become widespread

Howard R. Bowen

that higher education is conducted inefficiently, and effic-iency has been advanced as a major goal.

At the most global level, efficiency refers to theproper allocation of total resources to higher education sothat the benefit from the last dollar allocated will be asgreat as the benefit from the last dollar spent for other pur-poses such as private consumption, urban improvements,domestic transport, military operations, etc. The argumentis sometimes made that there are too many studentsenrolled, too large a research establishment, too manypublic services, etc., and that the growth of expendituresshould cease until balanced margins are restored.

A second concept is that the scope of the enterprisemay not be too great, but that higher education does notproduce as much return from its total expenditures as itmight. It is often alleged that it has surplus plant, the schoolyear is too short, teaching toads are too low, the pace is toslow, technology is backward, and managerial technique isslack. By internal tightening up, it could allegedly producea greater return at the same cost.

A third concept is that resources arc not allocatedproperly among the various branches of higher education,among the various fields of study, between undergraduateand graduate instruction, or between teaching and research.For example, too much attention is given to academicstudies and not enough to vocational studies, too manystudents are enrolled in expensive universities and notenough in community colleges, graduate study and researchare being overdone, too few M,D.'s are being produced, toomuch attention is being given to theoretical and esotericsubjects and not enough to practical and applied subjects,etc.

An inherent difficulty resides in these allegationsabout efficiency. Though the inputs can be measured bothphysically and in dollars, the outputs or outcomes arelargely non-measurable except through intuitive judgment.People of equal knowledge and integrity can reach quitedifferent opinions. Devices such as costbenefit analysisand program budgeting are not of much help when thebenefits are so hard to identify. Even comparative cost datafrom different institutions are hard to interpret becauseprograms are not the same and qualitative results may bequite different.

The efficiency question looks quite different tomany educators from what it does to outside legislators orbusinessmen. Educators see an educational establishmentof rapidly growing enrollments where society is continuallyloading on additional responsibilities, where faculty areworking long hours, where academic standards have beenrising and the richness of education has been increasingdespite inadequacy of funds, where the academic quality ofeducation and research is as high as anywhere in the world,where there is a perennial shortage of administrative and

9

GOALS AND FINANCING

non-adademic staff to get the job done, and where financialprecariousness is a way of life. The educator sees highereducation as being under-funded in relation to the enor-mous tasks to be accomplished and in relation to thereturns being yielded by expenditures for other private andpublic purposes. He sees higher education as comparingvery favorably in efficiency with many segments of privatebusiness, with the health services industry, with localgovernment, with federal agencies, with private founda-tions, and with the use of resources by private consumers.He sees his duty, in quest of efficiency, to seek more funds,not to cut back.

Educators are, of course, not the best judges of theirown efficiency, and some external check is reasonable.One possibility is simply to slow up the rate of increase ofexpenditures and thus force institutions to cut back theirexpenditures of lowest priority until the "fat" has clearlybeen squeezed out. Another possibility is for a coordinatingagency or some other public body to set standards ofexpenditure per student, perhaps on the basis of costanalysis, and to limit appropriations accordingly. The thirdis to rely on the price system to allow competition toforce institutions to hold down their costs to reasonablelevels for the kind of service they are rendering and tocharge high enough tuitions so that students will not utilizeeducational services in which they are not genuinelyinterested. One of the arguments often advanced forconverting higher education to the market system is thatefficiency would be promoted.

Minimizing the Scope of GovernmentA set of related goals are: (1) that government should

not engage in activities that can be conducted privately, (2)that public subsidies should be used sparingly, (3) thatpublic budgets should be balanced, and (4) that the pricesystem has special merit as a device for allocatingresources. These opinions cf course lead to the view thathigher education should be financed through tuitions andthat students should be financed through loans.

One important technical fact about loans for thefinancing of students is that loans can be financed throughthe private capital market and need not show up in publicbudgets. Except for this fact the global economic effects ofloans and grants are not very different. in either case, thefunds are likely to be supplied by financial institutions,and later repayment constitutes a mere transfer. Theeconomic cost of current higher education cannot be trans-ferred to the future,

EquityAn important goal is equity among individuals and

among social classes in the distribution of the benefits ofhigher education and in the distribution of the costs. In the

10

distribution of benefits, wide and easy access of all classesof people to higher education of all types is perhaps theonly requirement for equity. In the distribution of costs, theequity problem is more complex.

A common assumption is that costs should be distri-buted in some fashion between the student (and his family)and society (as represented by government and phil-anthropy). A reasonable basis of the division is often saidto be benefit. The student should pay an amount corres-ponding to his private benefit in the form of life enhance-ment and increased earning power, and society should payin proportion to social benefit. This formulation, however,presents difficulties because of differences of opinion notonly about the relative extent of private and social benefitsbut also about what should be included in the costs.

As the economist sees the costs, they include: (1)foregone income by reason of the student's being in collegeand not in the labor market, (2) incidental expenses ofcollege attendance, and (3) institutional costs of providingeducational services. 1 estimate the amount of these costs(annually per student) as follows:

Amount Percent

Foregone income $5,000 55%Incidental expenses 500 6Institutional costs 3,000 39

$9,000 100%

The question is: What proportion of this $9,000 should beborne by the student and what portion by society?

Suppose one were to assume (as seems reasonable tome) that two-thirds of the benefit were private to thestudent and one-third were social. Students on the average,then, should bear about $6,000 and the public about $3,000.The $6,000 would be met if the student gave his time to theeducational enterprise, paid for his incidental expenses,and contributed about $400 in tuition. Obviously theassumptions on which this result rests may not be univer-sally acceptable. But for what these assumptions may beworth, they suggest that a major part of institutional costsshould on grounds of equity be paid from appropriationsand philanthropy.

Those who discount the social benefits from highereducation and who reject the idea of foregone income willcome to quite different conclusions. Indeed, differences inthe interpretation of social benefits and foregone incomeunderlie much of the conflict of opinion about the finan-cing of higher education.

Equity in the distribution of costs also concerns fair-ness as between students who have affluent families toassist them and students who do not. If one assumes that

students are emancipated and that family income is not aconsideration, then all should be treated alike. Either loansor grants would be appropriate, On the other hand, if oneassumes that families when able are responsible to supporttheir children in college, then fairness may require that thegovernment and philanthropy act as surrogate parents forstudents whose families are not able to help. This wouldargue for grants rather than loans to needy students.

Still another equity issue concerns intergenerationalfairness. The American tradition of higher education hasencouraged low-income students to attend by means of lowtuitions, modest scholarships, and part-time work. Underthis system, generations of young men and women of low-income families have attended college without piling uplifelong indebtedness. We are now in the process of bring-ing another generation of young people, many of themminority origins, into the mainstream of American life viacollege. Is it fair to change the rules for this new group byrequiring them to go into debt?

Considerations of equity on the whole favor a systemof finance based on low tuitions coupled with grants ratherthan loans to low-income students.

Tempering inequality in Income DistributionA goal that is seldom mentioned directly but often

implied is to use the higher educational system as a vehiclefor reducing inequality in the distribution of income. Thedistribution of income is affected by higher education intwo ways.

Higher education can influence the distribution ofincome directly because it takes money from taxpayers andspends it on students. If the taxpayers are on the averagericher than students, income distribution will tend to beequalized; if the taxpayers are poorer than students,inequality will be accentuated. The conclusion is that eitherthe tax system should be made more progressive or tuitionsshould be raised.

A more fundamental influence of higher educationin income distribution derives from its long-run effect onrelative wages and salaries of various classes of the popula-tion. If the percentage of the population attending highereducation continues to grow, the supply of persons avail-able for professional, skilled, and white-colar occupationswill increase, and the supply available for blue-collar andunskilled occupations will decrease. As a result, the rela-tive compensation of the two groups will change, and thedistribution of income will become more equal. To achievethis result calls for a system of finance that encouragesaccess, namely low tuitions and grants to students,

Access and OpportunityOne of the most widely-held goals of American

higher education is that young people, regardless of theircircumstances, should have access to as much higher

Ho ward H. Bowen

education as they are qualified for and motivated to obtain.Today, this principle is being extended to adults of all agesand demands are being made for lifelong access. This goalof widest possible access is based on the ideas that accessis tantamount to opportunity, both cultural and economic,and that widespread higher education will foster economicgrowth through discovering talent, sorting people outaccording to their interests and abilities, and developingvocational skills. It rests also on the conviction, supportedby considerable recent evidence, that the innate ability tobenefit from college education is much more widely distri-buted than had previously been believed, and that newkinds of higher (or post-secondary) education could bedevised to accommodate a large proportion of the popula-tion. This goal of access accounts also for the recent drivefor "open admissions."

If access were the sole objective, it would call for asystem of finance with institutions funded primarily bypublic appropriation., and philanthropy and only second-arily by tuitions and with needy students funded primarilyby grants and only secondarily by loans. However, as analternative, access would not be curtailed very much ifinstitutions charged high tuitions but needy students wereassisted by correspondingly large grants.

Student FreedomIn recent years, the position of college students in

our society has changed markedly. They are increasinglyregarded as adults, they are given greater liberty in theirways of life, they vote, and paternalism on the part of bothparents and educational institutions is on the wane. One ofthe goals of higher education today is to maintain orincrease the newly-won freedom and independence ofstudents in their selection of colleges, in their choice ofcourses and educational programs, in the mode of instruc-tion and learning to which they are exposed, in theirpattern of living, and even in their relations with herparents,

This goal calls for the finance of institutions throughtuitions and the finance of students through grants or loansdirectly from government to the students, not via instit-utions. if students are the chief vehicle for bringing fundsto the colleges, and are free to select colleges of their choicethey will acquire substantial power over the educationalprocess. Colleges will be forced to be attentive to studentopinion and need.

The goal of student freedom also raises questionsabout the relation of students to their parents. In general,America has accepted the idea that parents are responsible,to the extent of their means, to finance the education oftheir children. But if the children are to be free and to beregarded as adults, they should at some point be emanci-pated from their parents. Should the time of emancipationbe at age. 18? Age 21? At graduation from college? At

11

GOALS AND FINANCING

completion of graduate and professional education?America has been equivocal on this subject of eman-

cipation, In Peneral, graduate fellowships and assistant-ships have been awarded without a parental maw, test,whereas undergraduate awards have usually involved ameans lest. Yet, the G.1. Bill, one of our most far-reachingand successful student aid programs assumed that theveterans were emancipated from their parents regardless ofage. In the past, when credit has been used to financecollege education, parents have usually been the respon-sbile borrowers. But recently, as the use of credit has beenincreasing, the students are often assuming the respon-sibility of repayments, A persuasive argument for thenewer loans of substantial amounts, long maturities, andcontingent repayment features is that such loans makepossible the emancipation of the student. Grants of equalamount would, of course, give them even greater freedom,but grants of equal amount are not likely to be forthcoming.

A more subtle aspect of student freedom is that thestudent should be free to choose his field of study withoutarbitrary admissions and retention standards, withoutarbitrary quota systems, and perhaps without differentialtuitions for different dis iplines. According to this view,the higher educational system should be responsive to theinformed choices of students and not be managed bymeans of arbitrary admission restrictions and quotasdesigned to regulate the flow of people to different fieldspresumably in accordance with estimates of future man-power requirements. The case of medicine is often cited asa field that presents arbitrary barriers to thousands of well-qualified men and women who would like to becomephysicians and who are needed in the health servicesystem. According to this concept of freedom, the highereducation system should be planned primarily with thedemand for places on the part of qualified students as theprincipal criterion rather than arbitrary estimates of prolessional associations, manpower planners. etc. A possiblefinancial implication of this point of view is that thecharging of different tuitions for different fields of studywould unduly influence choices because relatively smallsums of money might weigh heavily in the short run andproduce bad decisions for the long run. A contraryviewpoint is that tuitions should, in the interests ofefficiency, reflect the relative cost of different educationalprograms.

In general, student freedom probably calls for afinancial system of high tuitions coupled with generousloans available without a means test and free choice of dis-cipline without large fee differentials.

Academic FreedomA venerated goal of higher education is academic

freedom. This refers not alone to the right of professors toseek and speak the truth; it refers also to the power of

12

faculties individually and collectively to decide what linesof research and scholarship to pursue, what to teach, howto teach, what standards to maintain, what public serviceactivities are compatible with the main business of instruc-tion and scholarship, etc. The assumption has been that inall these matters professional judgment comes into play.

Obviously, professional prerogatives do not justifysocial irresponsibility, On the other hand, there is no sub-stitute for professional judgment in most educational andscholarly decisions. One of the inescapable problems ofacademic life is to achieve a reasonable balance betweensocial control and academic control such that the genuinelong-run interests of society are advanced. A major task ofcollege and university presidents is to bridge the distancebetween the academic interests and the interests of thesociety beyond the campus. There is no simple solution tothe problem, but the solution clearly does not lie indestroying academic freedom, in making the university agovernment bureau or market oriented enterprise. Topreserve the essentials of academic freedom, colleges anduniversities must have substantial unrestricted funds fromappropriations, gifts, endowment; they must also havediverse .-ources of support so that they are not beholden toany single interest or influence. To maintain significantinner direction, the college or university must be trusted bythe society; to maintain this trust while preservingacademic integrity is one of the continuing problems ofhigher educational policy and administration.

Knowledge and Learning as Values in Their Own RightIn most contemporary discussions, higher education is

viewed as a means to ulterior ends. However, learning onthe part of both students and faculty, the discovery anddissemination of knowledge, artistic creativity, and thepreservation and advancement of the culture can all beregarded as goods in themselves without reference to anyspecific instrumental purposes. Similarly, academic excel-lence can be regarded as intrinsically valuable. Moreover, ifthe experience of the past generation or two teaches usanything it is that knowledge gathered for its own sake, forthe mere value of knowing, often proves to be useful in themost unexpected ways.

This goal of knowledge for its own sake argues forsubstantial unrestricted funds for higher education to beused as the academic community decides. Indeed, highereducation is almost unique in our society as a place wherelearning and culture are pursued for their own sake. Oursociety should support at least one center where learning iscultivated in terms of its own intrinsic worth and notprimarily as a means to ulterior objectives.

Geographic Dispersion of Educational ResourcesEducational opportunity and resources should be

distributed geographically so that opportunities for people,

and also the cultural and economic influences of stronginstitutions of higher education, are available in all parts ofthe country. Without preventive efforts, the ablest faculty,the best students, and the richest resources will gravitate toa selected few institutions in restricted areas of the country.Adequate geographic dispersion calls for federal aid topoorer and sparsely-settled areas, and for institutionalcontrol of significant amounts of student aid funds.

Diversity and ProgressivenessA goal of higher education is to achieve a diversity of

institutions, programs, modes of instruction, and points ofview to accommodate students of varied interests andobjectives. A related goal is to encourage diversity in thesponsorship of institutions including federal government,state and local government, private non-profit organiza-tions, and private business corporations. In a time of rapidrelative growth of the public sector of higher education andincreasing political influence over the public institutions,the survival and prosperity of the private sector is of specialsignificance. The private sector, moreover, contributesmuch in le:;dership and quality to the higher educationalsystem.

The goal of diversity calls for either of two policies:(1) supplemental aid to students who attend private insti-tutions in order to compensate for the relatively high tui-tions in the private sector or (2) raising tuitions in publicinstitutions and thus narrowing the tuition gap.

Diversity and progressiveness also calls for substan-tial categorical aid for new institutions and new projects. Adanger in such categorical aid is that it will encouragephony or ill-advised innovation. Another risk is that gov-ernment and foundations will give temporary support forthe novel without properly encouraging the strengtheningand improvement of good features of the traditional. Solidchange is an evolutionary process. not keeping up with anendless succession of fads and gimmicks.

ConclusionsThe principal conclusion is that various goals, when

viewed separately, call for different modes of finance. Canone consolidate the various strands of the argument into acoherent system of finance? The answer is yes, but becauseof differences of outlook and priorities, one man's coher-ence can be another man's insanity. I shall try to presentthe conclusions I have reached about the financial patternthat derives from the fifteen goals.

Colleges and universities should have substantialunrestricted funds, preferably from diverse sources, in pur-suit of the goals of academic freedom, the advancement ofknowledge and culture for its own sake, the promotion ofnational economic growth, the solution of social problems,and the cultivation of citizenship. These unrestricted funds

Howard B. Bowen

might properly come in part from the federal governmentespecially for promotion of national economic growth

and solution of social problems.Federal and state governments should be able to

influence without dominating higher education bymeans of categorical grants for goals such as economicgrowth, solution of social problems, geographic dispersionof educational resources, diversity and progressiveness ofhigher education, and general responsiveness of highereducation to social needs.

A system of finance based on low tuitions coupledwith student aid emphasizing grants would contributetoward the goals of access, citizenship education, equity,and tempering inequality of income (in the long run).

On the other hand, a system of finance based onhigh tuitions coupled with student aid in the form of eithergrants or loans (with student aid independent of institu-tions) would work toward student freedom, diversity andprogressiveness, and improved income distribution in theshort run. If this system were extended so that colleges exuniversities were market-oriented in all or most of theiractivities, goals such as responsiveness, efficiency, andminimum scope of government would be favored.

To meet the goal of diversity and progressiveness,special provision would be needed to improve thecompetitiveness of the private sector, either: (1) tuitionequalization in the form of special grants to students tomeet the extra tuitions of private institutions, or (2) highertuitions in public institutions. I favor the first of these two,because I believe on other grounds that tuitions in publicinstitutions should be kept moderate.

To meet the 3oal of geographic dispersion of educa-tional resources, some substantial part of student aidshould be under the control of institutions,

In my judgment, no simple solution to the problemof financing higher education, based on only a few goalsand glossing over others, will serve. The only tenablesolution for the finance of institutions is a blend of unres-tricted funds and categorical grants from diverse public andprivate sources, moderate tuitions, and reasonable fees fornon-instructional services; the wisest solution for thefinance of low-income students is a combination of grantsand loans, with grants providing bare minima to assureopportunity and loans used as supplemental sources toassure flexibility. Such loans might be tong -term and haveincome contingent features.

My assessment of the goals leads me to questioncurrent tendencies toward converting colleges and univer-sities into market-oriented institutions without strong innerdirection, and equally to question current tendencies tofinance low income students primarily through loans. Theneed is for judicious balancing of many goals andbalancing the financial devices appropriate to these goals.

13

GOALS AND FINANCING

The system resulting from such balancing will never betidy or simple. It will never achieve any objective fully. Butneither will it destroy or handicap an educational systemthat has achieved first rank in the world in excellence andat the same time has extended higher educationalopportunity more widely than ever before in world history.

The same conclusion was reached by Marion B.Folsum, a distinguished businessman, former Secretary ofHealth, Education, and Welfare, and a long-time leader ofCED. He said :'

The financial support of higher education is apatchwork quilt. This support is drawn from vir-tually every known source . . . This patchwork quilt

. . is no jumble of confusion. Instead, it is a signifi-cantly complete list of the groups that form the broadbase of support for higher education in our society

. . If it is true that he who pays the piper calls thetune," the integrity of higher learning is ensured bythe fact that no one group can "call the tune." Thisbroad base of support ensures that our system will re-main free of a single, limiting educational creed. Andthis. in .a sense, is the genius of American education

that there is no single interest. no one creed ordogma, that might stifle the freedom and independ-ence we as people cherish,

This address was aided by a grant to Howard Bowen from the Committee for Economic Development. The content will also he pub-lished by the CED as The goals of Higher Education and Their Financial Implications" (forthcoming). It is printed here with permission fromthe Committee for Economic Development.

Marion B. Folsom, Who Should Pay for American Higher Education?" in Economics of higher Education, Selma 1. Mushkin (ed.} U.S.Department of Health, Education, and Welfare. OE 50027 Bulletin 1962 Number S. Washington: U.S. Government Printing Office, 1962, p. 195.

14

THE CHOICE OF A FUTURE

In following Robert Clark at the podium, my firstimpulse is to say,"Though I speak with the tongues of menand angels and have all knowledge so as to underst, nd allmysteries, because I have to follow Robert Clark, I'mdead." And Don, 1 appreciate what you said about me. Iwas wondering what you would find to say. Actually, I washoping that you would point out that my most significantattribute is that I'm so levelheaded. As you know inAlabama we have an operational definition of levelheaded-ness: that's a man whose tobacco juice runs evenly fromboth corners of his mouth.

An appropriate presidential address before the Asso-ciation for Institutional Research should contribute newinsights into the role and function of the profession so as toinspire the members with a clearer sense of direction intothe future. I had aspirations to deliver such an orationone that would not only vindicate the judgment of themembers who entrusted me with this high office, but thatwould also enshrine forever the memory of this time andplace in the hearts of my listeners. Having set this goal, Iprepared carefully.

My homework included a re-reading of the papersand addresses of all former presidents and such institu-tional saints as Dressel, Lins, Russell, and Brumbaugh. Atthat point, I understood the feelings of Charlie Brown onthe occasion when Lucy asked if he and Linus coulddescribe what they saw in the piles of cumulus clouds over-head. Linus reported that he saw the stoning of Stephen,Michelangelo in the Sistine Chapel, and Napoleon atWaterloo. Then Charlie Brown said, "I was going to say Isaw a goosey and a ducky, but I have changed my mind."

I also was required to make some change at least,insofar as the title of this address was concerned.Originally, it was called "An Inquiry into the Epistomo-logical Foundations of Institutional Research with SpecialReference to the Resolution of Certain TeleologicalAntinomies." However, the girls in Don Le long's officewould not print that in the program. Then I proposed"institutional Research and the Three B's Books,Buildings, and Bodies in the Garden of Learning. Willthey live happily ever after?". At that point, they suggestedthat my speech should be called "How to Meet Yesterday'sImperatives Tomorrow." Regardless of the title, I'm going to

Joseph T. Sutton, University of Alabama

talk about you and me, institutional research, and theAssociation for Institutional Research.

Can Institutional Research Make a Difference?A salient fact about colleges and universities today is

that the expectations they have held for the future havebeen grievously wounded. In the decades of the 60's and70's. we, in higher education, dreamed no little dreams. Itwas the best of years for physical planners. Communitycolleges sprang up over the country. Junior colleges andextension centers became senior colleges, and seniorcolleges became universities. Degree programs proliferated.Research flourished.

This pyramid of higher education has now begun toerode near the top and on the windward side. The presti-gious institutions and the privately supported colleges havebeen the more significantly affected, and nearly all institu-tions are sobered at their own prospects.

A second observation, made eloquently by Dr.Caruthers last night, and by Lyman Glenny at the 1971FJTUM in Denver, is that those who have traditionally beenrcsponsbile for the leadership of colleges and universitiesare less and less able to exercise control over them. HarlanCleveland, in his recent book, The Future Executive,contends that organizations are losing their hierarchialforms and that the organizations of the future "will beinterlaced webs of tension in which control is loose, powerdiffused, and centers of decision plural" Surely, this is anapt description of the current situation in higher educationgovernance.

Educational organizations with thwarted aspirations,increasing complexity, and with little control would appearto be less than ideal places to accomplish anything, butinstitutional research offices are found in such places orso I'm told. Indeed, some institutions have establishedoffices of institutional research because they have foundthemselves facing new and difficult problems. Perhaps thepromise of institutional research is brighter, the darker theinstitutional prospect.

The question for us today is: can institutionalresearch make a difference? Institutional research will notbe spared from the general muster to accountability. Myjudgment is that we can demonstrate that institutional

15

THE CHOICE OP A FUTURE

research is making a significant impact in understandinguniversity management, particularly fiscal management.However, the larger accountability of institutions to societyis neither financial nor managerial accountability. That isthe primary role of comptrollers, business officers, auditors,and examiners of accounts.

The institutions we serve are primarily accountablefor the effectiveness of their educational programs. Itfollows that the activities of the faculty, in their aggregate,comprise the prime functions of the college or university inaccomplishing this purpose. Increasingly, I fear, institutional research officers are seen by faculty as the residentSatanic apostle, emphasizing quantities rather than quali-ties, and efficiency rather than effectiveness. The existenceof institutional research on some campuses has become anadditional frustration to faculties.

The institutional research role is applicable to theentire university and its constituent programs. The Con-stitution of the Association reads that the major purposes ofthe Association shall be to benefit, assist, and advance re-search leading to improved understanding, planning, andoperation of institutions of higher education."

An institutional research officer should entertain thehypothesis that he or she can, by employing the tools andskills of his/her profession in a professionally competentway, contribute to the effectiveness of the faculty. Incomplex systems, control and direction are difficult to exer-cise. But institutional research officers understand systems,and it is this understanding that faculties need.

To assist in understanding the problem, consider for amoment the evolution of biological and social systems. Theprimitive unicellular protozoon, in its relatively simple or-ganization and structure, performed all the essential bio-logical functions assimilating food. eleminating wastes,moving about, sensing its environment, reproducing, etc.As multicellular forms appeared, so did cellular speciali-zation in function and differentiation in structure. Inspecializing. cells were able to do a more efficient jobofdigestion, or movement or coordinationbut in the processthese specialized cells lost some ability to do other func-tions and thereby became increasingly dependent upon themulticellular system of which they were a part.

Similarly, primitive man as a social organism had a verygeneralized social function in the family or tribal structure,but gradually roles and skills began to differentiate. Thesocial function of education eventually became a special-ized role for some members of the culture. In turn. educa-tional administration was differentiated from teaching, andin recent years, 1.R. has become differentiated from educa-tional administration and research. The first institutionalresearch was apparently done by college presidents. Someof them still do 1,R as a not yet fully specialized and dif-ferentiated function.

There are several points to this illustration. First, we give

16

up part of our freedom to make individual choices of afuture when we become members of a corporate group,society, tribe, college, or university. Interdependency withits consequent loss of control by the individual was, and is,the price that is paid for more efficient production andhigher living standards for the group.

Another indication from this brief look at the evolu-tion of systems is that not all institutions have evolved tothe same level of complexity and specialization. Some insti-tutional research officers do give major attention to theeducational programs of their institutions. If others of usfind ourselves becoming specialized in management, it is

time for another mitosis to occur so that the research func-tion as it applies to effectiveness of teaching will not belost.

Reviewing ProgressThe higher purpose of institutional research is then

to enable our institutions to continually review theirprogress, to project their many possible futures, and toselect among these, those that are both desirable and real-istic. There are many ways in which institutional researchcan fulfill this purpose.

Principally, I am speaking about education of thefaculty to the corporate reality of higher education as asystem. We recognize that we all retain significant emo-tional machinery that evolved in conditions where survivalwas at stake, and that this machinery responds to personalthreat with an overpowering need to do something about it.This need has many of us running off in several directionsat once. If there is to be an antidote to these individual"solutions" to higher education problems, institutional re-search must contribute to effective coordination and com-munication of the corporate purposes and goals of the insti-tution and the means whereby the institution achievesthese. 1 think this understanding is needed most severelytoday by the faculty. If, in our own sense of threat, weforget the threats perceived by others. especially the facul-ty, we have forgotten the unique role of institutional re-search in the institutional system.

The role of the Association has emerged from therole of institutional research in the institution. We need tocome together on occasions such as this to learn from oneanother. But we also need to be self-conscious about theAssociation so that we can corporately influence the futureof our profession and thereby our own. We have evolved toa significant level of complexity as an association and needto consider whether differentiation and specialization offunctions should occur. We need more effective feedbackmechanisms so we can develop a better information systemon our members and theiCaCtivities. We need to share thisinformation among ourselves and consider its implications.What are the implications to us all, for example, that onlyabout a third of our members give lull time to institutional

research activities, or that only about 20% of our memberscome from junior and community colleges? How is the roteof the Annual Forum changing, and how should it change?Should the Association develop a P. P. B.S. for its own plan-

Joseph T, Sutton

ning and budgeting purposes?The cobter's child goes unshod. Can the Association

contribute to the number of possible career futures fromwhich we as professionals may elect our own?

17

HIGHER EDUCATION UNDER STEADY-STATE CONDITIONS

Allan M. Cartter, Carnegie Commission on Higher Education

If I had known there were so many people here withlong memories, would not have brought the same speech Iused in 1905. Actually I remember that event very well. TheThree Village Inn in Stony Brook isn't up to the standardsof the Hotel Vancouver. It has lots of old world charm, butthe thing I remember was that the room was very dark. I amterribly near-sighted and can hardly see my notes on thepodium anyway, and they had a standing lamp over therewith a 40-watt bulb in it. I spent the evening standingunder the lamp. getting someone else to turn pages for me. Idid not have time to write this one, so I will not have asmuch trouble with turning pages.

The other thing that happened was that I had notbeen to Stony Brook before. That's why I accepted thatinvitation. 1 fell in love with it and six months later I

bought a summer house there, and for many years enjoyedgoing out and looking over their shoulder. I'd never been toVancouver before until this invitation came, so I'm going tocheck with Bob Clark later about real estate in this city,

It is true that I tried out that evening in 1965 for thefirst time publically, the faculty supply and demand projec-tions that I had been working on. I remember the last sen-tence of that speech. It was. "my advice to young facultymembers or graduate students is that 1970 will be a goodyear by which to have achieved tenure," Both AIR and Ihave aged a little bit since that time but that advice wasstill good. I think. I don't quite understand why 1 don'tpractice what I preach.. just two months ago I resigned mytenured professorship at NYU in Economics. It was afterthat meeting that some member of this Association sent mea little note with a quote from Sir Thomas Brown whichread:

Amuse not thyself about the riddles of futurethings. Study prophecies when they become his-tories. Eye well things past and present and letconjectural sagacities suffice for things to come.There are times when I wish 1 had listened to Sir

Thomas Brown's advice. But I think Fm the kind offrustrated institutional researcher myself. For the betterpart of ten years it was a kind of advocalion somethingthat one did on nights and weekends. This year has been akind of pleasurably traumatic experience for me for, partlyon the Guggenheim Research Fellowship and partly by the

18

Carnegie Commission on Higher Education, 1 now do from8:00 a.m. to 5:00 p.m. what I used to do by lamplight. I hopemy back cast friends notice that on the west coast at B a.m.to 5 p.m. we work a 40-hour week not including lunch. Ifind in switching from night to day time. that research onhigher education is rather different when you see it by day-light. There are certainly a lot more distractions, you don'tsquint quite as much, and when you get stymied on some-thing you can always pick up the telephone and find ananswer. You can't do that at one o'clock in the morning.

But even ominous projections aren't quite so depres-sing when you see them in the daylight. Last year Iattended on OECD conference in Paris on universitymanagement and one of our French colleagues who wasRector of one of the universities was making a speech. Theyhad simultaneous translations and suddenly the English-speaking community broke up with a roar of laughterwhich disrupted the meeting, We had to explain to thespeaker what the difficulty was. The French translationinto English came through as he was speaking about insti-tutional research and budget analysis as "some people liketo study futures. other people like to contemplate behinds."I want to suggest that that can be a kind of useful distin-ction. It's whether one's outlook is oriented to the past or tothe future. and I would pose that an applied science suchas Institutional Research, if it is really going to be effective.is of little value if it concentrates only on the past: if it doesnot provide some guidelines or guidance for the future. Thecorollary, of course, is, and my own experience well illus-trates this, that no futurist is worth much if he isn't familiarwith the past and does not try to test his projections of thefuture against the past experience.

My original battles and strong criticisms of theNational Education Association's studies on the Office ofEducation in the mid 1960's were largely because they builtmodels of the future that only a cursory analysis shouldhave indicated would not predict the present by glancingback to the antecedent. I have made a lot of errors inprojecting in looking at labor markets in the supply anddemand relationships, and I am embarrassed by a fewthings of mine in print, but it has almost always occurredbecause I have trusted the past too much, not too little. Butin today's uncertain world I think all of us are looking more

and more searchingly into the future and 'oo, are having tospeculate much more upon alternative courses of action.The presidents of universities and their trustees areincreasingly asking us what effect there is going to be onthe age. rank and tenure distribution depending ondifferent faculty personnel policies: what the effect is goingto be on enrollments of different tuition and scholarshippolicies: what is the voucher plan going to do to us: wouldthe Yale contingent loan program help or hurt our situa-tion: ho4is the changing ratio of men and women studentsgoing to affect curriculum, use of facilities, etc. So I thinkthat all of us are being forced to look more into the futureand to speculate more on alternatives. As I look to thefuture today. I think everything seems much murkier than itdid five years ago.

Let me cite a few areas where I think the past is nolonger a reliable prologue to the future.

College Enrolments

The obvious case is that of what's happening tocollege enrolments. As pessimistic as some critics felt myprojections of faculty needs were when I did them in themid 1960's, the situation now seems likely to be even worsethan when one first looked at it. Most of you are familiarwith parts of this evidence. Let me just quickly run througha few things that 1 think made it much more difficult todayto look ahead. The demographic factor: the birthratescontinue to decline in a rather frightening way and no oneknows when the situation is going to stabilize. Last winterfor the first time in history in the United States we droppedbelow the zero population growth fertility rate, When 1started making these projections in 1964 and 1965, we wereall forecasting the future based on Bureau of Census, Series'B' population projections which indicated that by the endof this century we would have an 18.21 age group of almost24 million people. In 19G9, when I went back and look anew look at it, we were talking about Series 'Ur projections,and instead of 24 million, it was only 19.1/2 million. InJanuary. 1972, in another piece that I did. I had to shift toSeries 'E' projections which looked as though they weregoing to be a little over 17 million college-age population bythe end of the century. By June, 1972, when I reached theCarnegie Commission, Gus Hagslrum has done the projec-tions there, and I went through again and we droppeddown to something about 7 or 8 percent below Series 'E'.And just two months ago the Census Bureau finailybrought out a new series 'F' which now predicts for the endof the century an 18-21 age group of 15-1/2 million. That'sexactly the number we have in 1973. That's a very differentkind of future than it looked as though we couldcontemplate 7 or 8 years ago. So over the next few years, upto about 1980, the college age group is going to grow byabout 1.1/3 million in the States. By 1988. however, it will

Allan M. Carter

have dropped 2,9 million down to about 14 million persons.I noticed in the New York Times on Tuesday that ArthurCampbell in the Population Research Center of NIH wasquoted as predicting a turn-around in fertility rates veryshortly, believing that the dramatic change in the last threeyears is largely the result of postponing having childrenrather than not having them at all, but that's no help to usbefore 1995. So certainly as one looks to the demographicfactors, things look much grimmer for those of us who arein the business of selling educational services that it didwhen we looked at it in the middle or late 1960's.

High School Completions

A second factor equally important is what's happen-ing to high school completions. All of my enrolment projec-lions and most others that I have seen recently are based onthe assumption that there is going to he a continuingincrease in the proportion of the age group that graduatesfrom high school. In my projections I have built in a 1%point per year annual rise from about 78% at the presenttime up to about 90% in the early 1980's. Now it appearsfrom recent figures that in 1970 there was only a .2 of 1%increase and in 1971 high school graduates dropped about1% . so the pool of potential students is already about 4%below the estimates I and the Office of Education andothers made only three years ago. I have not the foggiestidea of where we are going in high school experience:everyone seems to be puzzled by this reversal.

The third factor is what's happening to collegeenrolment rates, and again Iwo months ago, the CensusBureau came out with a report which I am sure all of yousaw, that the proportion of 18.19-year-old males in collegedropped from 44% in 1969 to 37.6% in 1972. For the 20-21 -year -olds, it went from 44.7% down to 36%. In the case ofwomen the 18-19-year-olds remained constant, 20.21-year-olds went up by about one percentage point. 13111 al least theproportion of the age group going to college in 1972 is nowabout 4 percentage points below what it was in 1969. I hadassumed in my projections. it would rise about 2.1/2% inthat period. The O.E. projections have built in about 3%rise, so we me far off the target in college enrolments.

The fourth factor, somewhat related, or at least addi-tional evidence, that while the percentage in college mayhave dropped. the question is what's happening to absolutenumbers. The Office of Education this year has done ahand tally of Fall 1972 enrolments and now we find thatfirst-time enrolment in degree credit courses was down1.4% for the United States in fall 1972 from the year before.There was a healthy increase in non-degree enrolments, butnot enough to really make up for that. First-year graduateenrolments in fall 1972 increased by only .06% They hadbeen going up 5 to 10% a year. So looking towards the next5 to 10 years. one could, on the basis of these somewhat

19

STEADYSTATE CONDITIONS

Shaky factors of the last Iwo or three years. if you tried toContemplate the worst, the high school graduates andcollege attendance rates would plateau out at their presentlevels and the graduate enrolments would stabilize if allof that happened, then all we could look forward to isabout an 8% increase in enrolments during the rest of the1970's and then, of course, a contraction fairly sharply inthe 1980's. If You compare this with the Office of Educationprojections that fulltime equivalent enrolment will go upby, I think, its M% in the 1970's, or even the CarnegieCommission assumptions that would rise by between 40and 45 percent. you can see that one can be very seriouslyshaken by this.

How much substance one should give to the eventsof 1970 to 1972 1 don't know. The change in the draft statusin Vietnam and temporary recession, unemployment, thesemay all just be very temporary things, but I think most ofthose who are watching these barometers rather carefullyare rather seriously shaken by it. Anyway, if in fact enrol-ments only went up by 8 to 10% over the rest of this decadeeven my most pessimistic projections of the demand forcollege faculty would look unrecognizably rosy by compari-son with the facts.

Lel me look very briefly at the supply side of theacademic labor market. Again, the rate of growth in Ph.D.output has been declining fairly sharply. The firstyeargraduate enrolments as a percentage of baccalaureatedegrees granted in the proceeding year actually peaked formen in 1966 and has dropped about 20% since then. Forwomen that percentage or B.A.'s going on to graduate workpeaked in 1968 and has dropped about 6% since then. Ifyou also look at the ratio of Ph.D.'s granted to first-year,firstlime enrolments in graduate school, on the averageabout 'five years earlier, you find that factor peaked for menabout 1970 and is now dropping fairly substantially. Forwomen is seems to have stablilized the last two or threeyears that's in effect a kind of completion rate ofdoctoral studies. Our latest estimates at the CarnegieCommission are that the 1972 Ph.D.'s. who numbered about34.000. that the number will grow the somewhere around 42or 43 thousand by 1980. That compares with the lastpublished O.E. projections of something like 68,700 sothere is some sharp adjustment going on on the supply sideof the market, too. In some fields the reversals are verydramatic. Physics is probably the best documented. In thecase of physics the percentage of freshmen who major inphysics in college has dropped from 1.8% of all freshmenin 1960 to .6% in 1972. So their share of the market hasdeclined down only 1/3 of what it was 12 years ago. Thefirst-year graduate students. in comparison with a numberof junior majors two years before that is more-or-less thepercentage that go on lo do graduate work has droppedfrom a high of 46% in the late 60's now down to 30 %, and

20

the number of Ph.D.'s granted in physics, since there areenough time lags built into the system, hit its high po,n1this last June at about 1,600. It looks as though i1 is going tobe down to only slightly over 800 by about 1977 or '78. Sothere are some fields that are adjusting very rapidly to thechange in job market conditions. There are other fieldswhere it seems to be happening in the reverse fashion. Themost perverse usually turns out to be in the field of educalion where for some strange reason I guess the tougher thejob market gets. the more important it seems to go on to gelone more degree to t.y and protect your position in yourcareer.Employment of Ph.D.'s

But strange things have been happening in theemployment market of Ph.D.'s. So strange, but againalthough I have spent a part of this year trying to look at it,I don't know how to decipher it Over the last five years thenumber of Ph.D,'s taking academic teaching positions, evenduring this period when we talked about the lough jobmarkets, the number taking teaching positions in highereducation has gone up forty percent. Two factors seem toaccount for this

One is an increase in the percentage of new facultyhired who already have the Ph.D. Institutions are hiringmany fewer who have lesser terminal degrees. The secondfactor is that there is much more filtering down going on asPh.D.'s now turn up much more frequently in institutionsof somewhat lesser repute. There are several interestingstudies that have been done by John Nyland that are veryrevealing on this. Comparing 1968 to 1971. he did a studyusing the- infamous Carter ratings trying to see whatproportion of Ph.D.'s in various fields took their firstteaching appointment in an institution that was of equal orof higher quality than the one from which they got theirPh.D. So if you got your degree from Harvard you had to behired by Harvard or you didn't fall into that figure. In thefield of economics in 1968, 20% of the new Ph.D.'s movedon to teaching jobs in institutions as good as the oneswhere they got their degree: 1971 it was down to 6% . Inmathematics it dropped from 29% to 9%: in geographyfrom 25% to 4%, and the overall figures were 19% and 8%.So there is a very sharp change in the pattern of types ofjobs that new Ph.D.'s are taking when they finish theirgraduate training. Obviously there is a bumping processgoing on and Ph.D.'s are bumping people with lesserdegrees or masters degrees or whatever in some of the slatecolleges and junior colleges.

Even more puzzling to me are the reports of the Na-tional Research Council that the number of new Ph.D.'staking appointments in the academic sector as postdocshas actually gone up by 60% in the last five years. In thissame period the number of post-doctoral fellowships andIraineeship awards has declined markedly, and yet

somehow. these people turn up as post-docs in the majoruniversities. What appears to be happening is a kind ofholding pattern that developed in a lot of the major univer-sities and they tend increasingly. now, to be hanging on totheir new Ph.D.'s if they don't find other attractive employ-ment immediately, and plugging them into ongoingresearch jobs in their own institution. In this way, they aresort of bumping what used to be the graduate student RA.and to some extent, even the teaching assistants. I find atBerkley, in some fields, they are even splitting post-doctoralappointments now. In Particle Physics most of the post-does are offered 50% appointments at $5500 or whateverinstead of full lime at $11,000, so increasingly there is akind of holding pattern that has developed over the last twoor three years. But I think it's fairly obvious that this canonly be a temporary phenomena. You can't go on enlargingthat reservoir continually; you can do it for a short period oftime largely at the expense of support of graduate studentsin the department. My personal belief is that the academiclabor market situation is going to get slowly but progres-sively worse in most fields less so in those which arequickly adjusted such as physics and the problem wouldbe less acute obviously in those fields where there is a fairamount of traditional employment in government andindustry or in the scientific fields, where I think employ.nient prospects have to rise again in the next several years.The really gloomy period. however, has never seemed tome to be the 1970's; it's always looked to be in the 1980's.So in one sense perhaps we should be thankful that thebudget constraints of the early 1970's have given us a lot offorewarning and that at least some adjustments now aretaking place within the system.

Steady-state ConditionsThis. looking from another point of view, poses a

kind of institutional problem that I suspect many of you arealso wrestling with and that is. how do we live in a worldthat's more akin to that of a stationary state without gettingtop-heavy in our faculty age and rank distribution. We havebeen playing with building models of the CarnegieCommission. faculty-flow models. Trying to see what wasgoing to happen if nothing else changes in the situation, butthese enrolment adjustments that we can see coming andwhat's going to happen to faculty hiring and the faculty ageand rank distribution. In 1970, 51% of college teachers wereunder the age of 40 and as we follow this model through,assuming everything else remains the same, the relativesalaries and retirements, that figure will drop from 51% to18% by 1990. If we have problems now with generation gapyou can see what it will be 20 years from now. In fact, thebiggest shock of my life is finding out that 78-112% of allcollege teachers are younger than I am. That did more tomake me feel ancient than anything I have ever done.Because there are so few faculty members, relatively. at the

Allan M, Carter

present time who are over age 50, it means that the impactof expansion of the last 10 and the next few years, meansthat the combined mortality and retirement rates of faculty,if there are no changes in retirement policies, which addedup to about 1.6% of the faculty either died or retirod in 1970

that factor actually is going to drop to about 1.2% in1985. So there is very little relief one can see until one getsalmost to the turn of the century in terms of much higherretirement rates. I started out about two years ago when Iwas getting alarmed by this to talk about lowering thecommon age retirement for faculty members. In fact. I thinkmy last act at New York University before leaving was todrop the retirement age from 68 down to 65. Then i got outof town the next week! I have now come to the conclusionthat really doesn't have much impact on the total system. Itmay help the individual institution but very little reliefwould occur from merely lowering the retirement age.Obviously that drop from 51% to 18% of the faculty thatare under age 40 is not going to work that way: the marketwill make some adjustments. There are some changestaking place in tenure policy, some changes in retirementpolicy, there are going to be some adjustments in relativesalaries between academic and non-academic employmentand the situation is going to be a little fluid. But it'sinteresting to try to built your first model and see whatwould happen if nothing else changed. Now we are goingback, and trying to see what happens if you build invarious assumptions on these adjustments.

Let me turn briefly to two other areas where the pastprovides very few clues for the future. I believe that in thelast several years there are many signs indicating thatpublic policy is beginning to shift around more and more tothe view that one should support students rather than justinstitutions. The imposition of tuition al the University ofCalifornia in 1970 may have signalled the beginning of this.The higher education amendments in 1972, 1 think, if oneanalyzes them carefully, are a clear indication that Federalpolicy is shifting in this direction and when Bob Kibbeethrows in the towel. then we will really know it's a trend.Every year the city university is threatened with moving toa tuition fee; they have not done so yel. The forthcomingCarnegie Commission report, due out at the end of June, isgoing to urge a narrowing of the tuition gap over the nextdecade. the gap between public and private institutionsplus a sitbstanlial increase in financial aids to students. Igather there is a Committee on Economic Developmentreport that's going to be out very shortly with somewhatsimilar types of recommendations and there are many statereports like the Governor's task force in New York thathave come up with recommendations of that sort.Increasingly, it seems to me this is becoming more andmore the fairly common popular view as the direction inwhich we should go. II is seen as a way to provide adequateresources to make universal access to higher education a

21

STEADY-STATE CONDITIONS

reality while at the same time trying to preserve diversity ofinstitutions. The California State Scholarship Commissionin its study of how college students pay for the financing ofhigher education done in 1972 is indicative in this regard.We at the Carnegie Commission have combined some ofthe California data with our estimates of institutionalsubsidies and have come up with the following estimates.

In the public sector (hit's taking the two-yourand four-year colleges all together the students andfamilies in the highest income group (that's $18,000 andover in the State Scholarship Commission study), studentsand families in that highest income group contributed only51% of the total monetary outlay as we have defined it,which is all of the institutional costs of education. the sub-sistence cost of the student minus whatever student aidthey had. So 51% was contributed from those in the highestincome group. Students and families in the lowest incomegroup contributed 35% . In the private sector, those per-centages are 54% and 31%. Bit there seems to he increasingpressure, I think, to try to create somewhat greater equityand many people argue that there ought to be greater dis-parity between them. Perhaps those in the highest groupought to move to 65 or 75% contribution and those in thelowest income group, if we are really going to do away withthe financial obstacles to attendance, ought to drop down to15% or 10%. But I believe there is increasing opinion thatpublic subsidies need to be redirected to be of greater assis-tance to the low and lower middle groups. I think that isgoing to make a considerable change in the 1970's on manyof us.

A Learning SocietyFinally. let me turn to one other area which is

rapidly becoming part of the public consciousness andcould have a marked impact. Over the past decade we haveseen a change in nomenclature for education to personsbeyond the normal college age. We have spoken of adulteducation, which became continuing education, and nowincreasingly you hear comments about recurrent educa-tion, the life-time learning society. etc. I think it's morethan just words. The new civil rights emphasis in thiscountry, I think, has given us somewhat greater concern foradults who missed out when they were 18 to 21 and neverdid get any advantage of post-secondary education andmost particularly so when you look at studies like that ofBob Hartman at the Brookings Institution that indicates that40% of the public subsidies that go to higher education arepaid for in taxes by people who are nonattenders thatis, who did not get any benefit directly from it. Similarly,somewhat in the civil rights vein, the increasing re-entry ofolder women to the labor force. many of them wanting to

22

go back and get refresher courses or to complete education.In addition to that, the rising problems of technologicalobsolesence is making us more aware of the need forretraining at various times in one's life. The manpowertraining problems that have emerged because of the riseand fall of industries like aerospace, the increasing practiceof young people of stopping out in higher education andthen at 25 or 30 wanting to come back to complete all ofthese things are contributing to a new awareness and a newkind of demand for higher education on the part of adults,Last year the OECD held a very interesting conference onnew patterns of life in adult years. The chief argument wasthat study, work. and leisure all ought to be seen as part ofa continuum and that separating out financing retirementsone way, and education in another way, didn't make muchsense. The recommendations for the OECD membercountries were that they ought to explore moving to somekind of common social insurance program which wouldrecognize that there are trade-offs between lifetime educa-tion, periods of leisure, adjustments in retirement age, etc.The International Labour Organization is holding a bigconference in Geneva this summer on the subject of paideducational leaves. The new programs introduced withinthe last two years in France and West Germany havesuddenly stimulated a great deal of interest. There is risinginterest among the trade unions in this country and it'squite possible that this will bring more and moreawareness and more and more pressures for doingsomething to help finance other than traditional 18.21education. The Commission on Non-traditional Studies, thereport of the new Commission on Continuing Educationthat Father Hesburg chaired, the report due in August orSeptember by the Carnegie Commission. all of these thingsare adding much more interest and much more concernwith the factor.

The likely next major development. I think. isdevising some type of contributory plan to finance adulteducation. If so, it could revolutionize the educationalstructure of higher education. At the present time onlyabout 25% of these activities take place in what we think ofas colleges or universities. A lot of it is out in civicorganizations and proprietary schools. etc.

I have catalogued a number of areas which many ofyou either now do or will shortly deal with in one respector another. They are all areas in which guidelines seemunclear at the present time and where extrapolation is

inadequate. II makes the job more challenging if moredifficult. The only thing I think that we know for certaintoday, and perhaps we should be thankful for this onething, is that famous punch line of the comedian Mort Sahl,"the future lies ahead."

THE HAZARDS OF PLANNING PREDICTING PUBLIC POLICY

For those of us who have watched with alarm andbewilderment the shifting breezes of Federal involvementin education during the past decade, the juxtaposition ofthe words pi,u;ning and public policy in the same sentenceapproximates a contradiction in terms. Yet both publicinterest and the public investment in higher education havebeen growing rapidly, facts which by themselves force usto deal serious! with this topic.

Predicti, g public policy as it relates to or affectshigher education is not a pursuit which many educatorshave mastered, in fact it is not something to which many ofus have given much thought. But since we must think aboutit now and in the future it behooves us to do so in somestructured way and hopefully with as much informationand analyses as we can muster,

What I intend to say on this topic today is notintended to provide you with either a method or a modelwith which to approach this task. Rather I shall attempt todemonstrate the influence of public policy in four broadareas of institutional concern and to suggest how they willaffect radically the operation of colleges and universities inthe future. Niuch of what I shall refer to could have beenpredicted by people assigned to systematically scan thepolitical horizons even though my own awareness of them Ireadily attribute to hindsight rather than foresight.

Given the importance of public policy to theoperation of colleges and universities in the future (which Ihope to demonstrate) the syllogism is easily constructed bywhich predicting public policy becomes essential to soundplanning and thereby a matter of concern to those repre-sented here this morning.

For purposes of this discussion I am defining publicpolicy to mean any consensus of ideas and attitudes whichhas been converted into an operational principle throughlegislation, administrative regulations or court decisions.From this it can be readily recognized that public policymay be something different from popular belief and that itcan operate at several political levels. Individual slates oreven localities can establish policies that ai the time are notwidely accepted. Some of the issues I shall discuss are rotyet broad national policies although I am convinced thatthey soon will be.

Robert J. Kibbee, The City University of New York

Let me begin by saying a few words about thereasons for the increasing application of public policy tohigher education. The litany is familiar to most of you but itserves to provide a proper context for what follows.

GrowthThe reason for the growing interest in higher educa-

tion is not difficult to understand. It is a function of its size,its pervasiveness, and its cost. In what is really little morethan one generation, higher education has projected itselfinto the mainstream of our domestic concerns largelythrough its phenomenal Arth, In the United States betterthan six out of ten hig!, school graduates now go on tocollege and the percentage is still growing. This means thattoday virtually every family has had some immediate ex-perience with college either directly or through the acti-vities of someone close to them. Today our colleges anduniversities enroll about 9 million students, employ nearlyone million professionals and an almost equal number ofnon-professionals. Total expenditures for higher educationnow exceed 30 billion dollars annually, roughly two-thirdsof which are tax dollars. Thus there is a high level ofinterest in higher education because it touches virtuallyeveryone either in their personal family lives or in theirpocketbooks two of the most sensitive of all humannerve endings.

The corollary to this growth is more subtle but in thelong run, will be more significant and brings me to the firstarea of public policy I wish to discuss.

A generation ago when 1 was in college no more thanone high school graduate out of seven went on to collegeand a much smaller percentage of those who entered highschool graduated than do today. Thus the percentage ofcollege graduates in the total cohort of young men 21 25 atthe beginning of the second world war was very small,indeed. Except for the major professions such as law andmedicine it was quite possible, even probable, that youngpeople without a college education could make it into themainstream of the nation's economic life and rise through itas far as their talents would take them. Not so today, andeven less so tomorrow. As the percentage of those withsome college education rises to be a major fraction of their

23

PLANNING PREDICTING

age group, college becomes a giant sieve through v.hichsociety sorts out those who can succeed from those whomust accept what is left. When this happens, and it is al-ready happening, the opportunity to attend college is nolonger a luxury but a necessity and when it is a necessity itwill rapidly become a right. The growing demands ofminorities for an opportunity to attend college is based atleast in part upon their perception of a selection processbased upon the college experience. These demands willgrow as the noose tightens.

This seems to be an appropriate point at which tosuggest a few principles to reflect upon.

The first prindple is that the ability to excludepeople from an opportunity in our society is in inverse pro-portion to the percentage of the population that is includedand to the importance of that opportunity to participation infuture benefits of the society.

The second principle is that the ability to excludepeople from a benefit of our society is in inverse proportionto the public tax dollars invested in providing that benefit.

These are hardly in a class with Newton's Laws but Ibelieve they state with some accuracy what has alreadybegun to happen in higher education and to what I believewill happen at an accelerated pace in the future.

Continued ExpansionThe proliferation of community colleges throughout

the nation, the growing attacks upon the traditional admis-sions processes at colleges and universities: the initiationand expansion of state scholarship programs and the "openadmissions" policy adopted at my own institution arc all, inwhole or in part, reflections of the operations of theseprinciples. I cannot imagine that the pressure will subside.Those who must predict what the future holds would dowell to consider the probabilities of universally availablehigher education and how such a public policy might affecttheir institution. Even private institutions which might tooreadily relegate the implementation of such a policy toothers should pause to reflect on the real possibility that tothe extent that they do or might receive public subsidy theymay be expected to participate in the solution of this publicproblem.

It also seems safe to predict that within a relativelyshort time the loose relationships that have characterizedthe internal operations of the campus are going to becomeincreasingly formalized, This is a euphemistic way ofsaying that the relationships among the several elements ofthe college campus are rapidly going to be governed byunion agreements arrived at through collective bargaining.Although many may wish to debate whether this is good orbad. I find such a discussion, if you will pardon a pun,academic. The movement has started, it is growing and itwill not be stopped. Those who think it evil may resist it.those who find it sensible may embrace it but in the long

24

run it will come and we will learn to live with it.On many campuses today. large segments of our

non-professional employees are unionized and the numberincreases daily. Within the last few years the NLRB hasproclaimed its jurisdiction over large and middle-sizedcolleges and universities and each year state legislatures areextending to teachers the right to collective bargaining andeven the right to strike. Today, faculties at about tenpercent of higher education institutions in the United Statesare represented by a recognized bargaining agent. By thetime a new decade dawns I feel certain that virtually allpublic institutions and most large private institutions willbe dealing with faculty unions of one sort or another. But itwill not end there. The movement will be slower indeveloping but within the time span we are talking aboutstudents will be in this game too. It is highly unlikely thatthey will acquiesce for long to being whipsawed betweenthe demands of the faculty and the desires of theadministration. The era of effective student unions is atleast within sight.

As the formalization of relationships develops it willraise many questions about traditional educationalpractices that will have to be resolved. Among these are therole of the faculty in curriculum development. methods ofselecting, evaluating and promoting faculty. the wholequestion of tenure: and a hest of other matters that havebeen part of the mythology of higher education.

Moreover, it will alter significantly the staffingrequirements of institutions, record keeping, hiringprocedures and legal services, all of which will affect theallocation of resources.

A New AccountabilityThis brings us to our third point. In our recent

history no matter of public policy has affected us moredirectly or more pervasively than in the area of personalhuman rights. It is true that we are participating in thisinstance in a major national. if not international.phenomenon. But it is significant that previously heldexemptions are being withdrawn gradually from collegesand universities either by legislative action or judicialmandates.

We are, of course, most familiar with the anti-dis-crimination edicts and the affirmative action programs ofthe Federal government but in many areas the surveillanceof human rights is multileveled. The City University ofNew York. for example. in addition to satisfying Federalrequirement may be challenged through the uniongrievance procedures. the City Human Rights Commissionand the State Division of Human Rights. Moreover, anyindividual may take his complaint to each of these tribunalsconsecutively or to several of them concurrently. Theimplications of this for university operations can be readilyimagined.

But those aspects of human rights which are relatedto discrimination arc only part of the picture', Recent curtdecisions have ranged over a much broader canvasincluding due process in the handling of disciplinary casesof both students and faculty, proper procedures for facultyappointments and nonreappointments, residency require-ments for students acid the control of student publications.These decisions have, in effect, applied general publicpolicy to campus operations and in the processsignificantly altered traditional relationships and opera.tional methods. I am certain that we can expect this trendto continue both by expanding constraints on areas alreadytouched and by taking up operational areas not yetaffected. Surely the possible effects of this process cannotbe excluded from our planning if we are at all concernedabout future reality.

My final example and one which relates mostdirectly to the size and cost of higher education, l shall tryto encompass with the general term "accountability." Inone sense the concept of accountability transcends thegeneral area of public policy since it has captured theimagination not only of legislators and governors but of themany nonpublic benefactors of our institutions includingfoundations. alumni, corporations and even students andtheir parents.

For years all educational institutions have beenaware of the necessity for fiscal accountability in the sensethat they have been required to assure their constituentsthrough independent audits that money received wasproperly accounted for. Many public institutions have foryears been subjected to varieties of pre-audits of expendi-tures which have often promoted the green eyeshadedenizens of budget offices to faculty-like status. In morerecent times many public institutions have had to justifyboth new buildings and additional staffing requests byextensive analyses of space utilization and class size. And,of course. in virtually all institutions. the administrationand faculty have always been accountable to their trustees.

But am speaking here of a different kind of ac-countability. It can translate ultimately into the simpleconcept of "more bang for the buck" and reflects a concernfor both the efficiency and the effectiveness of the educa-tional enterprise. Discussions of the new concept ofaccountability are peppered with terms borrowed fromindustrial management-productiviy-cost-benefit .and soforth but the meaning is clear. The various supportingconstituencies of higher education are concerned aboutboth the process and the product. The reality of thisconcern cannot be judged by the vagueness of the solutionsthat are being suggested.

Unfortunately, we are woefully unprepared torespond to these concerns in terms that are convincing. Theproduct of the educational process is elusive enough in

Robert J. Kibbee

itself. When one tries to measure the value-added througheducation the task appears staggering. Yet if we areunwillinp, to try to frame the discussion in our own termsthere are others who will do it for us. probably in ways thatare less than satisfactory.

Since efficiency is always easier to measure thaneffectiveness I fear that those outside the academy willsettle for measures of quantity rather than of quality andwill focus on teaching rather than on learning. If we are tomove the discussion to a different plane we must moverapidly to address these questions in our own house.

It is clear to me from these and other less obviousmanifestations that questions of public policy areimpinging with accelerating frequency and with greaterpervasiveness on e'er colleges and universities. The fewinstances that I have mentioned here this morning havealready had a profound effect upon the operations of someinstitutions and, I believe, will spread to all or almost allbefore this decade is done.

Changing Public PolicyNew thrusts undoubtedly lurk on the horizon and

will soon be upon us. Moreover, we have ample evidencethat public policy does change and we suspect thatanalyses would show us that certain kinds of publicpolicies shift more rapidly than others.

As we plan for the future it is essential that weattempt to factor this increasingly important considerationinto our future thinking. Flow we go about this process isextremely important and those of your who are here todayhave a vital role to play in that process.

I have recently read a small monograph whichdetailed the planning dynamics of a major university over aperiod of almost ten years. Since I am fairly familiar withboth the institution, the individuals involved and the milieuin which the planning too place, t have reflected a gooddeal on what I read. 1 have concluded that given the normalplanning process some very intelligent men devised aprogram based on decisions which at the time they madethem were almost uniformly correct. Unfortunately, theyhad assumed that the world they were operating in wouldbe the same tomorrow as it was today but more so. Theworld changed and many of the correct decisions becamedisastrously wrong. A large element of the difficulty arosefrom changes in public policy. Whether or not the shifts inpublic policy might have been perceived had this aspect ofthe environment undergone careful ccrutiny one will neverknow, The fact is that it was not done and that thedecisions were made in the kind of unsuspecting euphoriathat grips small investors in the stock market just before thebears take over.

Since I have proposed here that public policy is animportant ingredient in the planning process, that it will

25

PLANNING PREDICTING

become increasingly important in the future and that,therefore, it is essential that we make an effort to predictwhat that policy is likely to he, it seems only fair that I saya few words about how this might he done.

Most obvious is the necessity that if any systematiceffort is to be made it must be somebody's resonsibility todo it. The Office of Institutional Research or the Office ofPlanning seem logical candidates for this assignment,There are several places where one might start. Sincepublic policy tends to change slowly and grows through anaccumulation of individual instances, it might be wellto follow carefully the literature of higher education toperceive which ideas seem to develop a following. Theannual reports of foundations, particularly the larger ones,are another source of information not only for what theysay but for an analysis of those ideas or concerns whichattract support. Certainly some way of monitoring thepublic policy decisions that are made in those states whichtend to be in the forefront of educational change willsuggest clues to what might become general policy in thefuture. The public statements of governmental officialsgovernor, legislators, agency heads are another source of

26

useful information, particularly if one is alert to ideas thattend to reoccur with increasing frequency. Court decisionsaffecting higher education are already monitored byscholars who have identified trends that might usefully beconsidered by those who must plan for the future.

What we are talking about is a form of educationalintelligence work that gleans information from manysources, evaluates it, pieces it together into a meaningfulmosaic and attempts to determine how it might affect whatone wishes to do in the future. It is not an exact science butrather the application of analytic intelligence to informa-tion. The process does not guarantee sound decisions but itdoes insure that a vitally important component of futureplanning will be consciously addressed.

Planning is both an essential and a hazardous under-taking. Predicting the future we have left largely toastrologers and mystics. But if planning is to be more thanwishing we must make an effort to predict what the worldwill be like. If predicting is to be more than guessing wemust do it as carefully and as systematically as we can.

wish you luck,

PROGRAM EVALUATION: APPROACHES AND PROCEDURES

There is an increasing concern with program evalu-ation in higher education. This concern stems from a de-mand for new academic program to meet the changingneeds of college students and from an insistence that pro-grams be properly managed and systematically judged,

The Carnegie Commission on Higher Education(1972) has issued a strong appeal for the development ofacademic programs that would serve the needs of studentswho have not traditionally sought higher education andwhose motives and interests differ greatly from previousstudents. The clear implication of that appeal is thatstudents are changing and academic programs must bechanged accordingly. The Commission on NontraditionalStudy (1973) makes an even stronger appeal for academicprograms that would serve students rather than institutions.permit a degree of flexibility and an adaptiveness that hasoften been advocated but seldom realized, and help realizea total learning society that would place no limits on theefforts of its citizens to acquire the skills and knowledgethey seek.

Other converging events that spur the concern withprogram evaluation are the federal legislation of the sixtiesthat made specific requirements for evaluation and thecurrent dominance of economic and engineering conceptsin educational thought and discussion. The Elementary andSecondary Education Act of 1965 (ESEA) gave an impetusto evaluation that may indeed have a greater impact oneducation than the act itself. The allocation of massivesums for the improvement of education did carry a man-date for insuring that the money was well spent. Theensuing disillusionment with the massive programs fundedby ESEA underscored the need for better methods, skills,and techniques for the evaluation of educational programs.To quote Malcolm Provus:

Perhaps before we can build effective new pro-grams. we must establish creative new ways tomonitor and eventually judge the effectiveness ofsuch programs. This capacity to evaluate programsmust ultimately depend upon a managementtheory that utilizes pertinent. reliable informationas the basis for administrative decisions (1969. p.243).

The dominance of economic and engineering con-

Cameron Fincher, University of Georgia

cepts has emphasized the shift of conceptual focus frominput and process variables to an assessment of educationaloutcomes. The overplayed theme of accountability is basedon a concern with efficiency in the utilization of personnel,resources, and materials in the educational enterprise. Asmuch as educators may disclaim the appropriateness ofeconomic and engineering concepts for academic programs,they should be nonetheless cognizant that such conceptsare the dominant ones at the present time.

The purpose of this paper is to examine the currentapproaches that are being made to program evaluation andto consider some of the implications for academic programsin higher education. The effort is predicated on the basisthat program evaluation is a management imperative forhigher education and vitally affects the role and function ofinstitutional research in our colleges and universities. Theevaluation of academic programs in the past has not beenviewed within the larger framework of policy analysis andhas not been a major concern of institutional research assuch. Not only have past efforts been spasmodic and piece-meat, but the evaluation of academic programs has notbeen well grounded and has often produced results thatconclict with professional experience and judgment. TheColeman Report (1966), the study by the RAND Corporationfor the President's Commission on School Finance (1972),and the much publicized Jencks study (1972) are indicativeof an increasingly persuasive argument that runs counter toour better hopes and expectations for education. In brief,these studies raise a number of critical questions for publiceducation as it is traditionally valued in American society.They leave us wondering most seriously where the diffi-culty lies in the educational programs as implied, in themethodology of evaluation as it is imperfectly developed, orin the choice of criteria as they are opportunisticallyemployed. The serious implications of such studies stronglysuggest that they cannot be easily ignored and theycannot be accepted at face value.

The dilemma posed by the numerous efforts at pro-gram evaluation in the past several years is. to say the least,embarassing it would seem most advisable then to ex-amine recent trends in program evaluation, to understandcarefully the conceptual and methodological difficultiesthat are involved, and to seek a better conceptual grasp of

27

PROGRAM EVALUATION: APPROACHES

the nature and functions of evaluation in general.

Recent Trends in Educational EvaluationThere is a well-known but highly embarrassing sus-

picion that most efforts to evaluate human productivity, ef-fectiveness, or well-being wind up on a pessimistic note.This is true not only of programs in education but also ofevaluative efforts in the fields of social welfare, penal cor-rections, public health, counseling, psychotherapy, rehabili-tation, public housing, and religious conversion. Whentraditional methods of experimental or qtl-si-experimentalinvestigation have been applied to the problems of humanbetterment, correction, enlightenment, or uplift, the resultshave not been as encouraging as proponents of theprogram, technique, procedure, cure, or remedy had hopedin the beginning. Whatever human nature is, it does notappear as malleable, flexible, or plastic as reformers, re-vivalists, and crusaders proclaim. Change in human be-havior is not easily assessed in systematic, quantitative, orexperimental fashion and an inherent resistance to changehas been postulated more than once to avoid the full pessi-mism of research findings.

Perhaps the encouraging feature of the most recentwave of pessimism is the growing recognition that thescientific methods developed for investigation of physicaland biological phenomena in a laboratory setting are notadequate for the evaluation of programs seeking to im-prove, correct, or benefit human beings. It is not the wrong-ness of scientific method that is underscored by this out-look but rather the limitations of narrowly construedmethods in dealing with social, cultural, and psychologicalcomplexities.

A changing Notion of EvaluationAs in so many other changing trends in education,

the National Society for the Study of Education was quickto address one of its yearbooks to a consideration of theissue and problems involved. Edited by Ralph Tyler (1969)the 1969 yearbook addresses itself to the changes takingplace in educational evaluation. New needs, conditions,knowledge, and techniques are said to foster new concepts,procedures, and instruments of evaluation. Massive finan-cial support was first given by the federal government tothe development of new courses in science and mathe-matics, then to the social problems of poverty and discrimi-nation. Each produced in turn a demand for sound datathat would guide public decision and policy. Coupled withthe rapid increase in technological instrumentation, federalsupport of education spurred a concern for evaluation thatcould not be met by traditional test theory and practices.An adequate appraisal of new conditions and the applica-tion of new knowledge about education as such wasbelieved to necessitate new developments in evaluation.

28

Merwin (1969) emphasized that change had beenprevalent in educational evaluation since its inception inthe 19th century but changing concepts of evaluation werecalled for by the new educational efforts and technicaldevelopments, Bloom (1969) found one of the major theo-retical issues of evaluation to be the degree of specificationof educational objectives and outcomes, The basic problemis making the desired outcomes of learning sufficiently ex-plicit for communication to teachers, technicians, students,and the evaluators themselves.There is also the problem ofevaluating the non-specified outcomes of instruction,judging the effect of evaluation on the learning processitself, and then assessing the usage of evaluation results. Heposits that evaluation is more likely to have effect when (1)important and relatively stable characteristics aremeasured, (2) the data are used for important decisions andmajor disjunctions, (3) the results become part of the publicrecord, and (4) they are used to judge the effectiveness ofteachers, schools, or school systems.

Stake and Denny (1969) make the cogent observation thatevaluation is not "a search for cause and effect, aninventory of present status, or a prediction of future success(p. 370)." It is something of all three but only if it contri-butes to understanding substance, function, and worth.Evaluation may have many purposes but should alwaysattempt "to describe something and to indicate its per-ceived merits and shortcomings." They also point out thatthe successful application of evaluation methodology is de-pendent upon people and that one of the most urgent needsis for more effective recruitment and training of evaluators,This need is not supplied by graduate programs in tests-and-measurements or in research-design-and-analysis. Torealize the potential of evaluation, different concepts andtechniques are needed, Other needs are better means ofrepresenting educational goals and priorities, reliable tech-niques for assessing instruction and learning materials, andsimple instruments for gathering opinions and value judg-ments. Needed most is the development of formal pro-cedures for processing evaluative information and drawinginferences that warrant the confidence that is placed inthem.

A Sharpening FocusThe American Educational Research Association's

recurrent interest in educational evaluation is indicated bythe April 1970 issue of Review of Educational Researchand the monograph series it has launched on the subject ofcurriculum evaluation. The growing concern for evaluationtheory, research, and practice is reflected quite strongly inthe issue of RER and several of the conceptual shifts in roleof evaluation are pinpointed quite well. Cohen (1970!stressed the radical change in the purpose of evaluation asthe result of the national thrust against poverty and dis-

crimination. The large-scale programs of the federal govern-ment meant that educators were now involved in social ac-tion programs and not merely education as they were ac-customed to it. This meant further that the whole businessof evaluation research had been thrusted in the politics.This turn of events was a source of confusion to many edu-cators who were not prepared for the shift in national goatsand expectations.

The AERA Committee on Curriculum Evaluationwas established in 1965 and commissioned to developguidelines for the development and revision of educationalcurricula. The Committee has proceeded on the belief thatcontemporary testing and measurement procedures areinadequate and that special observation, data-gathering,and decision-making techniques are required. To en-courage the development of a more adequate theory andrationale for the use of such techniques, the Committee hasissued several monographs in the series that open up dis-cussion in a commendable fashion.

Scriven's (190) distinction between formative andsummative evaluation has been widely adopted. Hedescribes the former as the efforts of educators to improvecurricula during their development and the latter as thefinal or conclusive evaluation that would be made aftertermination of a course, program. or project. Formativeevaluation is, of course. intermediate in the sense that itpermits adjustments, modifications, and improvementswhile the program or project is still in a slate ofdevelopment. The focus is on deficiencies and weaknessesthat should be eliminated or on strengths that should beemphasized. In summative evaluation the focus is onjudgments or over-all conclusions that would assess theterminal worth or value of the program. The finality orconclusiveness of the evaluation is, needless to say. still amatter of relative degree but the implication is thatsummative evaluation borders cia the final word that maybe had on the subject under investigation.

Scriven's notion of formative and summativeevaluation is comparable in many ways to the distinctionsthat Cronbach and Suppes (1969) made elsewhere betweendecision-oriented and conclusion-oriented research. Tradi-tional methods of science are said by the latter to bedirected to the derivation of general conclusions and not tothe solution of immediate or practical problems.Conclusion-oriented research is directed by the researcher'sown commitments and intellectual interests. He raisesquestions as the research proceeds and is expected tofollow leads that may prove more promising than his initialor primary hypothesis. 11 is expected that his final resultswill contribute to the large body of scientific knowledge inthe particular area in which he is working.

Decisionoriented research is conducted for thespecific purpose of facilitating decisions that must be made

Cameron Pincher

in the immediate future. The purpose of the research is notdetermined so much by the researcher's interests and com-mitments as it is by the needs and dictates of persons ad-ministratively responsible for the decision. Timeconstraints and other restrictions are placed on theresearcher's efforts and he is not free to follow up side-leads or by-paths that might prove more interesting than hisoriginal charge. When the research is completed. theresearcher may or may not be free to publish his findingsin a scientific or professional journal. His audience willmore likely be the administrative, legislative, or publicauthority for which he did the research and hisresponsibility will not be limited to publication of hisresearch findings. To the contrary, he may spend monthsexplaining his findings to groups of critics, adversaries, andothers who view his research findings with suspicion, ifnot outright hostility.

Evaluation and Decision MakingThe most ambitious approach to program evaluation

has been made by the Phi Delta Kapp Study Committee onEvaluation (1971). This Committee makes a direct attack onthe notion that evaluation is "a healthy science, operatingfrom a base of well-established theory and methodology,and with obvious benefit for all (p. 314)." Evaluation isdefined by the Committee as "the process of delineating.obtaining, and providing useful information for judgingdecision alternatives (p. 40)."

The central concept in evaluation, as presented bythe Committee, is decision. The decision process is said toconsist of the phases of awareness, design, choice, andaction. Decisions are said to be made in a setting that iseither metamorphic. homeostatic, incremental, or neomobi-listic. Decision models that are applicable for the decisionsettings arc the synoptic, disjointed incremental, andplanned change models. The types of decisions called formay be either planning. structuring, implementing. orrecycling decisions.

To mesh the concept of decision with educational evalua-tion, four basic kinds of evaluation are identified ascontext. input, process, and product. A flow chart, depict-ing the basic elements of decision settings, models, andtypes, are believed by the committee to present a totalevaluation model for use by those who would approachevaluation systematically.. The extent to which they actuallysynthesize a new definition and methodology of evaluationis not immediately ohvious. Yet, given the criteria set bythe Committee, the problems involved in developing anadequate theory of evaluation, and the unresolved issuesthat remain, the PDK Committee has written a book thatanyone taking evaluation seriously should not ignore. Theyhave definitely designed wheels that need not be re-invented. Yet, the complexity of the evaluation model and

29

PROGRAM EVALUATION: APPROACHES

the ambitiousness of their efforts to treat all forms ofevaluation may have resulted in a mote detailed model thanmost persons concerned with evaluation will desire. Thereis a criterion of fruitfulness that remains to be fully tested.

Evaluation ResearchThe emergence of program evaluation as a research

specialty is depicted more fully by the work of EdwardSuchman (1967), Rossi and Williams (1972), and CarolWeiss (1972). Others could be added to the list, but thethree will suffice.

Weiss' test is based on the theme that evaluationuses the tools and methods of social research in a contextof social action where traditional research methods oftengel bogged down. As social research, evaluation demandsan imposing array of talent. Not only are research skills re-quired but also political, social. and relational skills thatenable the evaluation researcher to utilize his skills for thepurpose of improving the rationality of social policy. Thepurpose of evaluation research, as stated by Weiss, is to"measure the effects of a program against the goals it setsout to accomplish as a means of contributing to subsequentdecision making about the program and improving futureprogramming (p. 4)."

Rossi and Williams believe evaluation research is aninput to policy analysis. Its purpose is to provide evidenceof the effectiveness of existing program and proposed al-ternatives as a basic guide to decision making. As such, itinvolves both the evluation of progam outcomes and fieldexperiments that would test the merit of new ideas withprogrammatic implications. Both forms of research shouldbe distinguished from onsite monitoring and should helpimprove the social policy process. Rossi and Williams leaveno doubts that policy-oriented re,-earch is indeed political.Not only are the problems met, Jdo logical but they areorganizational and bureaucratic. They state explicity thatthe role of the policy analyst and the evaluation researcheris basically forensic. II is their task to gather evidence thatwill persuade and influence those who must make deci-sions concerning the program under study. The continu-ance, modification, or cancellation of the program maydepend directly on their skill in doing so.

As Harold Orlans (1973) has pointed out, the termprogram evaluation occurred only rearely in research litera-ture until the domestic programs of the federal governmentproliferated in the sixties. Now there seems to be a concensus among social scientists familiar with governmentalaffairs that a much needed form of research is one thatwould evaluate the effectiveness of governmentalprograms.

Yet. Orlans is dubious that social scientists are eitherwilling or able to provide the kind of evaluations that are soobviously needed. There are basic incompatibilities be-

30

tween research as it has been taught in most graduateschools and evaluation research as it is needed in the arenaof public policy. Orlans writes that, "evaluation also faces aset of inter-linked moral and practical problems that impairits objectivity, insight, and candor (p. 123)." The successfulevaluation researcher will be hard put to supply the adap-tability, ingenuity, and resourcefulness that can produce aninformed, careful, and relatively impartial research report.The motives to criticize whatever findings or conclusionshe reaches will be numerous and uninhibited.

A Conceptual Framework for Program EvaluationAn emerging research specially may or may not be

reflected in the conceptual shifts seen in educational evalu-ation by the various committees of NSSE, AERA, and PDK.The evidence does suggest, however, that a more sophisti-cated form of program evaluation is necessary and that itmay not be readily provided by persons trained in conven-tional research methodology, The difficulties stem from dif-ferences in approach, procedure, and expectations; they areboth conceptual and methodological in nature.

The conceptual shifts that are involved in the chang-ing demands placed upon educational evaluation may besummarized as: (1) an increasing emphasis on the use oftests and measurements for feedback and direct improve-ment as opposed to selection and placement of students, (2)an increasing advocacy of developmental, applied,practical, problem-solving research as opposed to basic,longterm, theory-based research, (3) a lessening of interestin conventional experimental research as opposed to re-search that is administratively or action oriented, and (4) a

growing interest in decision-oriented research as opposedto conclusion-oriented research. For the evaluation ofacademic programs, these shifts would suggest that aheavier emphasis should be placed on formative, asopposed to summative, evaluation and that the results ofevaluation studies will be used increasingly for policy deci-sions that will shape or determine program substance,form, and content. The extent to which evaluation researchwill be cast in the mold of operations research or insti-tutional research remains to be seen, Also in question is theextent to which evaluation researchers can maintainscientific objectivity and detachment as these values aretraditionally conceived. The degree of active involvementthat is implied in some evaluation efforts requires a signi-ficant alteration of the researcher's professional role as ithas previously been viewed.

An adequate conceptual framework in which to viewprogram evaluation would seem most desirable. The modeldeveloped by the PDK Committee is highly suggestive andshould help conceptualize the problems and issues ofevaluation in a more meaningful fashion, There is aliability of intricacy. however, that may prevent the model

from having the influence it deserves. Persons asked toconduct evaluation studies will continue to be drawn fromtraditional fields and may find the model to be needlesslycomplex for their immediate purposes. As a result, manybeneficial effects will not be gained.

To provide something of a conceptual framework forprogram evaluation in higher education, a provisional efforthas been made in Figure 1. No effort is made to treat theframework as an operational model or to imply that it haselaborate structural features. it is strictly a functional repre-sentation of the classes of variables that are involved tovarying degrees in evaluation efforts. The intent is toprovide a perspective from which to view the difficulties,problems, and issues of evaluation. Many efforts at evalu-ation research would seem to suffer from too narrow afocus on either the process being investigated or on thetangible products or outcomes of that process. The con-ceptual framework should help demonstrate that there is anoptimal distance from which to view most educationalprograms and projects: there is risk involved in standingeither too close or too far back. It is possible both tomagnify process variables out of proportion to their im-portance and to blur their features to such an extent that

Cameron Pincher

the process acquires blackbox status.The conceptual framework or design has been

borrowed from Brunswik's writings on probabilistic func-tionalism and should suggest the advantages of bothprobabilism and functional analysis. A probabilistic ap-proach should help avoid a narrowly construed deter-minism that sets the researcher in search of oneto-one re-lationships between the several classes of variables. Thetime perspective should help disabuse any researcher of thenotion that his efforts should serve to link remote events toultimate effects. These two classes of variables areprovided primarily for boundary purposes and should helpdi;;courage the excessive ambitiousness that some evalua-tion studies have been susceptible to in the past. The func-tional analysis of distal events and eventual outcomesshould give any researcher as much as he could hope tohandle.

It should be obvious from the conceptual design that achoice of variables is crucial to the process of evaluation.Proximal events will be identified as those variables thathave been traditionally classified as independent or experi-mental variables while concurrent events will be quicklyrecognized as extraneous variables that must be controlled

Figure 1.

A CONCEPTUAL FRAMEWORK FOR PROGRAM EVALUATION

(adapted from E. Brunswik, "The conceptual focus of some psychological -0,1ems," (ournal of Unified Science, 1939, 8,36.49)

REMOTE EVENTS DISTAN EVENTSPROXIMALEVENTS

Mill

CONCURRENT IMMEDIATEEVENTS RESULTS

I is

#

MOM'11111.f 11111111MIN

3

I'llSITUATIONAL

FACTORS

EVENTUALOUTCOMES ULTIMATE EFFECTS

31

PROGRAM EVALUATION: APPROACHES

if functional relationships are to be clearly established.Situational factors may be seen as a special class of vari-ables in that they have an anchoring effect on the decisionsor conclusions that may be derived from the evaluation,This is to say that virtually all evaluation results will havesome degree of specificity that cannot be generalized toother situations and circumstances.

Immediate results will be seen as comparable to thedependent or behavioral variables that are identified inconventional experimental designs. These will be the vari-ables in which the researcher will be most often interestedbecause they are more accessible, more tangible, and moreeasily communicated to others in decision making posi-tions. If, however, there is a serious concern with eventualoutcomes, the evaluation study must be properly executedto that end. The distinction between immediate results andeventual outcomes will depend, of course, on the re-searcher's own time perspective. The distinctions arebroadly implied, however. and if eventual outcomes are tobe treated as dependent variables. the rationale and pro-cedures for doing so should be specified. On the otherhand, if immediate results are to be treated as proxies foreventual outcomes, the reasons for doing so should beexplicit.

Distinctions between types. levels, and forms ofevaluation are not inherent in the conceptual design butshould be facilitated to some extent by its consideration.An obvious inference is that all classes of variables neednot be included in a particular evaluation study. If the con-ceptual framework provides any help at all, it shouldconvey the foolishness of efforts or notions that we canevaluate academic programs in a total or comprehensivemanner. The effectiveness of any evaluation effort willalways depend on the researcher's ability to identify,

analyze, and interpret those variables that bear mostdirectly on the process, outcomes, or events he has chosento study. The choice of variables, therefore, should alwaysbe explicit and as consistent as possible with the overallobjectives of the evaluation study. For example, the choiceof variables in organizing and implementing a system ofprogram evaluation should reflect the difference between amonitoring system and a means for individual certification.just as testing practices have been dominated by reliabilitytheory as a means of providing highly stable measures ofindividual performance, so can evaluation efforts bederailed by the mistaken notion that the purpose is tocertify individual competence. The two approaches aredifferent and require different techniques and procedures.

In closing, it is well to mention that the conceptualframework should incorporate without undue difficulty themajor advantages and implications of the schema, models.and approaches discussed previously. The primary insis-tence of the conceptual design is that systematic efforts atprogram evaluation should constitute a functional analysisof the variables that are involved and that the nature of therelationships derived or established will be probabilisticrather than deterministic. It is believed that such a con-ceptual approach can help avoid many of the difficultiesthat have been previously involved in our efforts to evalu-ate academic programs. The complexities of most programsin higher education and the subtlety of influential variableswould strongly suggest that as urgent ns evaluation fa, itwill not be easy to conduct the kind of evaluation studiesthat are needed. The methodological difficulties are stag-gering but they are not insurmountable. There is everyreason to believe that our success or failure in programevaluation will depend primarily on our solution of theconceptual problems that are involved.

Averch, Harvey A.. Stephen I. Carroll, Theodore S. Donaldson, Herbert I. Kiesting, and john Pincus. How Effective Is Schooling? ACritical Review and Synthesis of Research Findings. Santa Monica, Calif.: RAND Corp., 1972.

Bloom. Benjamin S. Some Theoretical Issues Relating to Educational Evaluation." In Ralph W. Tyler (Ed.), Educational Evaluation:New Roses. New Means. Chicago: University of Chicago Press. 1969. pp. 26-50.

Campbell. Donald T. and Julian C. Stanley. Experimental and Quasi-Experimental Designs for Research. Chicago: Rand McNally.1963. (Reprinted from Handbook of Research on Teaching).

Carnegie Commission on Higher Education. Reform on Compus: Changing Students. Changing Academic Programs. New York:McGraw Fl ill, 1972.

Cohen, David K. "Politics and Research: Evaluation of Social Action Programs in Education." Review of Educational Research, 1970,40. 213-238.

Coleman, lames S., et at Equality of Educational Opportunity. Washington. D.C.: U.S. Government Printing Office, 1966.

Commission on Nontraditional Study. Design for Dive. sit y. San Francisco: fosseyBass. 1973.

Cronbach. Lee I. "Course Improvement Through Evaluation." Teacher's College Record. 1963. 64, 672-682.

32

Cameron Finder

Cronbach, Lee J. and Patrick Suppes (Eds.) Research for Tomorrow's Schools: Disciplined Inquiry for Education. New York: NationalAcademy of Education, 1969.

Denny, Terry (Issue Editor). ''Educational Evaluation." Review of Educational Research, 1970, 40(2 }.

Jencks, Christopher, et al. Inequality: A Reassassment of the Effect of Family and Schooling In America New York: Basic Books, 1972.

Merwin, Jack C. "Historical Review of Changing Concepts of Evaluation." In Ralph W. Tyler (Ed.), Educational Evaluation; New Roles,New Means. Chicago: University of Chicago Press, 1969. pp. 6-25.

Orlans, Harold. Contracting for Knowledge. San Francisco: Jossey-Bass, 1973.

Provus, Malcolm. "Evaluation of Ongoing Programs in the Public School System." In Ralph W, Tyler (Ed.), Educational Evaluation:New Roles, New Means. Chicago: University of Chicago Press. 1969, pp. 242-283.

Rossi, Peter H. and Walter Williams. Evaluating Social Programs: Theory, Practice, and Politics. New York: Seminar Press, 1972.

Scriven, Michael. The Methodology of Evaluation." Perspectives of Curriculum Evaluation, Chicago: Rand McNally, 1967, pp. 39-83.

Stake, Robert E. and Terry Denny. "Needed Concepts and Techniques for Utilizing More Fully the Potential of Evaluation." In Ralph W.Tyler (Ed.). Educational Evaluation: New Roles, New Means. Chicago: University of Chicago Press, 1969, pp. 370-390.

Stufflebeam, Daniel L.. Walter J. Foley, William J. Cephart. Egon C. Cuba. Robert L. Hammond, Howard 0. Merriman, and Malcolm M,Provus (PDK National Study Committee on Evaluation). Educational Evaluation and Decision Making. Itasca, HI: F.E. Peacock, 1971.

Suchman, Edward A. Evaluative Research: Principles and Practice In Public Service and Social Action Programs. New Yark: RussellSage Foundation, 1967.

Tyler, Ralph W.. Robert M. Gagne. and Michael Scriven. Perspectives of Curriculum Evaluation. Chicago: Rand McNally, 1967.

Tyler, Ralph W. (Ed.) Educational Evaluation: New Roles, New Means, The Sixty-eighth Yearbook of the National Society for ihe Studyof Education, Part 11, Chicago: University of Chicago Press, 1969.

Weiss, Carrol IL . . 1972.

33

BETTER PROGRAM EVALUATION:

HOW CAN WE MEET THE CONCERNS OF TODAY WHICH WILL BECOME THE IMPERATIVES OFTOMORROW?

For the past five years I have been coordinating anexhaustive literature review on the nonintellective correl-ates of different kinds of college program outputs whichwill hereinafter be referred to as "college success." Thisstudy was inaugurated in the summer of 1967 by Commis-sion IX (Assessment for Student Development) of theAmerican College Personnel Association, and has beencontinued under the sponsorship and funding of TheAmerican College Testing Program (ACT).

What one person considers success may not besuccess for another, and this is definitely true on thecollege campus. For example, some students merely wantto persist while others would consider it a failure if theydid not graduate with honors. Some consider their primarypurpose to be preparing for a job while others are con-cerned about developing their social skills, developing aphilosophy of life, finding a suitable marriage partner, etc.In addition to college success as seen through the eyes ofthe student, college success can be defined from the per-spective of the colleges and their officials, the govern-mental agencies, society in general, and interested personssuch as parents or relatives, etc.

The initial phase of the project involved searchingthe Psychological Atrsracts back through 1957 for researcharticles dealing with anything tha; might be consideredcollege success. These references and pertinent informationwere recorded on specially per cared evaluation sheets.Over 2,000 references were thus ide 'ed, after which thesheets were sorted into criterion categories and then intosubcategories.

Once some college success categories and the foci ofthe study had been ascertained, a thorough search of theliterature was initiated. Between 5 and 6 thousand differentpertinent references of published literature were found,and this does not include unpublished reports or disserta-tions. Furthermore, this figure does not include publishedreferences prior to 1963 in the grades, persistence, and aca-demic learning areas because a large number of publishedreviews of literature were found which seemed to ade-quately summarize the literature for those areas to 1963.

One of she noteworthy results of the project on col-lege success wt s the development of a criterion classifica-tion system with broad categories and subcategories of

34

Oscar T. Lenning, American College Testing Program

success as defined by various publics. As mentionedpreviously, the categories of college success were, in amanner of speaking, empirically derived. For each criterionarea specified, a number of research studies relating thecriteria to other variables were found. Some of thesestudies actually attempted to predict the criterion whileothers were concerned only with trying to provide insightsand to broaden the level of understanding of the criterion.

About 400 of the research studies were selected forannotation. The annotations and references (grouped ac-cording to criterion and predictor categories) have been or-ganized into two monographs, which are expected to bepublished by the end of 1973.

As the many studies were critically reviewed to seewhich should be annotated, a number of observationsabout research methodologies and pitfalls were observedwhich should be of interest to institutional researchers.These observations and concerns seemed to he of problemsthat also presently hinder effective college program evalua-tion, and they are the topic of this paper.

I am not trying to equate ''educational research" with"program evaluation" in this paper. For example, evalua-tion must often utilize rough indicators which would notqualify for designation as formal research measures. andmuch of program evaluation does not involve use ofstatistical tests. Evaluation should be a continuous processthat starts with the beginning development of a programand continues throughout the life of the program. H is anongoing process that is continually feeding back informa-tion to the program coordinators and others. Furthermore, itchanges and refocuses as the program is modified inresponse to such feedback. An educational research projecton the other hand tends to be strictly summative in naturewith pre-established research hypotheses as guidelines forthe research study.

What I have been talking about is "formative evalua-tion." After the program is operational, much of theformative evaluation data is pulled together into "sum -mative evaluation- form. and usually some additionaloverall evaluations of the program outputs are conducted.This summative evaluation is in effect local educationalresearch or institutional research, and many of thepublished studies of college outputs were of this type.

When program evaluation reaches the summative stage, it ishelpful to be able to compare your evaluation experiencesand results with those obtained for simila', program at othercolleges. The discussion throughout the remainder of thepaper focuses on this aispect of program evaluation,

The Research Pitfalls NotedIn a delightful address in 1965 to Division 14 of the

American Psychological Association, Marvin D. Dunnette(1966) outlined six "fads. fashions. and folderol inpsychology" which he felt summarized what was wrongwith the profession. Several of his points are very muchapplicable to the research on college success which wasreviewed. His first heading was titled ''The Pets We Keep"or alternately "What Was Good Enough for Daddy Is GoodEnough for Me." He was talking about a prematurecommitment to a particular theory or research method. Theanalytic methods used in a number of the studies raisedquestions of appropriateness. Perhaps the reason some ofthese studies used inappropriate methods was that theresearchers had a "pet" method they tended to always useor perhaps many of them were unfamiliar with other moreappropriate methods and did not have access to astatistician.

Something that may be contributing much to theproblem of using inappropriate methods is our system ofgraduate education. It is questionable whether the majorityof graduate schools in this country are providing anadequate background in research design and methodology.Secondly, there is a tendency for particular "schools" ofthought and methodology to predominate in a university.Futhermore, it is probable that some graduate students areoverly influenced by the preferences and biases of theirmajor professor.

Dunnette's second heading was entitled ''The NamesWe Love" oe ''What's New Under the Sun." What he wastalking about here are the imprecise definitions and thevague constructs prevalent in various areas of educationand psychology. He says:

Perhaps the most serious effect of the game is thetendency to apply new names in psychologicalresearch widely and uncritically before sufficientwork has been done to specify the degree ofgenerality or specificity of the "trait.' being dealtwith. Examples of this are numerous anxiety.test-taking anxiety, rigidity, social desirability,creativity, acquiscence, social intelligence, and soon ad infinitum. (p. 345)

Dunnette's next title was "The Fun We Have" or"Tennis Anyone?" In this fad, the researcher gets so caughtup with computer capabilities or testing null hypothesesthat he forgets the real problem and also perhaps forgets topersonally look at the data itself. Dunnette believes this isprimarily responsible for what he calls the "little studies"

Oscar T, Leaning

or the "little papers" of psychology, which were found tobe so prevalent in the research on college success. Actually,small studies at single institutions can make importantcontributions to the institution, as most directors of Institu-tional research can attest. In addition, such studies canmake real contributions to the field if they deal with partsof important problems and if they can be synthesized andintegrated. The negative reaction of Dunnette is not againstsuch "little studies." but rather against those little studieswhere the research is poorly designed, poorly documented.and exploring something of little worth. Furthermore, theemphasis throughout the sixties on "number of researchpublications" as an important variable for faculty evalua-tion has been a central part of the problem for educationalresearchers outside of the Institutional Research Office,Because of this emphasis, some researchers have been moreconcerned with adding to their publication list and withmerely going through the motions of research than withmaking contributions for the good of their institutions andfor the good of our young people and of our society,

A final heading of Dunnette's that should bementioned is The Secrets We Keep" or "Dear God. PleaseDon't Tell Anyone." He reported a paper by Wolins (1964),who wrote to 37 authors asking them for the raw data onwhich journal articles were based. Of the 32 authors reply-ing. 21 no longer had their raw data. Wolins did reanalyseson the seven sets of data he was able to obtain and foundthat three of the studies had gross errors which changed theoutcome of the study. Dunnette includes in this categ.),ythe following types of problems, all of which abounded inthe reports on studies of college success: statisticaldifference tests reported without their correspondingmeans, SDs, and the correlations between the twovariables; experimenter biasing factors; incomplete descrip-tions of methodology; failure to carry out or report cross-validation studies; and failure to carry out or report reptica-tion studies. Dunnette also reported, and this may or maynot be extensive, that some researchers have the improperpractice of routinely dropping subjects from their analyses.

It was mentioned that summative program evalua-tion did not always need to involve statistical tests.However, when such tests are conducted all necessary datafor replication and comparison by others should beincluded in the study report, Similarly, the description ofthe total project should be comprehensive enough thatothers reading it have the complete picture and can easilyevaluate the adequacy of the evaluation study.

After reading Dunnette's fascinating account, is itany wonder that syntheses of a large number of studies inthe literature are difficult to accomplish? It seemed almostimpossible to really synthesize the findings from all thelittle studies reported in the literature because of suchfactors as differences in sampling procedures, differencesin criterion definition, and lack of documentation con-

35

BETTER PROGRAM EVALUATION

corning the design and the results, Through use of a setevaluation form. we have been able to gather as muchcomparable data as possible from each study. However.many studies did not report all essential data. In addition.there were also a number of other serious research pitfallsnoted in many of the studies reviewed during the project.

Many of the studies used improper control groups ordid not use any control group when one was clearly calledfor For example, no attempt was made in almost all of thestudies of the institution's overall impact on student changeto examine change in comparable nonstudents during thesame interval of time for which change was beingexamined for college students. Most studies that foundpersonality change and other change in college studentshave merely assumed that the college experience was afactor in bringing about the changes, e.g., the observedchanges in autonomy, authoritarianism, dogmatism, andindependence. Similarly man's studies of the impact ofparticular programs only looked at the change itself, anddid not compare it to the change taking place in students atthis college who were outside of the program. They justassumed that the program was involved in the change.

Concerning studies of change. many of theresearchers made no attempt to control for student inputcharacteristics. For example, it is common for change to berelated to aptitude, and so aptitude is one variable thatshould often be controlled.

There are a number of ways to control for inputs.Examples of simple experimental methods are matching orstratification. However, sophisticated statistical methods ofcontrol such as analysis of covariance may be called for insome cases. Because of a large sample size matchingprocedures may be impractical, and sometimes matching orstratification do not give the degree of control desired,Stratification gives only gross control on the variables ofconcern. Secondly, the desire may be to use the intactgroups for which differences are to be examined,Futhermore, analysis of variance requirements of pro-portionality of cell frequencies may prohibit the use ofstratifying. and matching may distort the distribution of oneor more of the groups being compared. in addition, it maybe desired to control on a number of groups and in somecases it becomes desirable to later control on unanticipatedvariables. For research strategies in studying change orcollege impact. see Harris (1963) and Feldman (1970).

Many of the student samples in single-campusstudies were quite small and they often appeared to be just"grab" samples with no real attempt to have representative-ness. Furthermore, many of the small samples on whichone-tailed t tests were used probably did not meet the"normal distribution" requirements for such a test, and atwo-tailed t test or nonparametric tests should have beenused instead. Different statistical procedures have differentdistribution requirements. For example, when group

38

comparisons are to be made, the relative homogeneity ofthe groups becomes important, Researchers all too oftencompletely ignore such considerations.

Many studies did not explore the two sexesseparately. From other studies. it was apparent that therewere important sex differences on many of the variablesbeing explored. To demonstrate the seriousness of this,suppose you are studying change and it happens that bothsexes change a large amount, but in opposite directions, Ifyou look only at change for the total group. you willerroneously conclude from your results that no change hastaken place.

II would also be desirable in program evaluation toexamine special subgroups of students within the programto see if the program has differential effects on studentswith different characteristics, For example, in the literatureon leacher evaluation, Domino (1970) found thatconforming students tend to do well in a class taught in astructured manner and poorly in a class taught in anunstructured manner, with the converse being true forstudents with an independent orientation. Most of thepublished studies only examined diverse groups ofstudents such as freshman or a psychology class, and didnot explore subgroups.

Another methodological problem was the use ofinvalidated instruments. Some studies used locallyconstructed instruments that were poorly designed andappeared to never have been field tested in any way. Nomatter what your location or clientele. it should bpossibleto try out an evaluation questionnaire ahead of time on afew people rather than relying entirely on "arm chair"evaluation of the questionnaires. However, even some ofthe widely used and accepted personality tests orinventories as well as other published instruments havequestionable validity because of the nature of theconstructs being studied. Care should be taken in choosingsuch standardized instruments to make sure that they areappropriate measures of what you wish to examine for theparticular program you wish to evaluate. The specific goalsof the program are crucial in making such decisions.

A final problem noted was the "helter skeller" waythe social science professions have of disseminatingresearch results and of applying them to practical situationsso that they will eventually have a genuine impact on thestudent and his college and on society. Not only is there alack of coordination in communicating research results, butmuch of the research is quite old by the time it ispublished. A number of institutional researchers havemade good progress in overcoming some of the cornmunication problems within the institution. One reason for this isthat institutional research people tend to be much moreoriented to research application than are academicresearchers. Futhermore. most of their research is designedto provide direct decisionmaking assistance. Because of

this research emphasis, many institutional researchersmake conscientious attempts to summarize their data intocharts and other forms that are easily understood by non-statistical people; to write up separate. short, and eyecatching reports directly aimed at specific college officials ontheir campus; to emphasize those results which have themost practical significance; to make use of media such asthe faculty newsletter and the campus newspaper for com-munication when appropriate, etc. Even so, such people arecontinually trying to improve their intra institutional com-munication methods.

Institutional researchers have in the past been at adisadvantage when it came to communication amonginstitutions. Research journals were generally notinterested in decision-making studies limited to onecampus, and institutional research conferences were fewand far between until the decade of the sixties was wellunderway. One problem, of course, was that until the1960's institutional research was universally carried on bymany departments and people throughout each campus,and there were no central institutional research offices tocoordinate and assist in such research efforts, and to focuson campus-wide research concerns. Hopefully, the ERICsystem will continue to make progress in reforming ourresearch inter-institutional communication system. butsomething more is needed as suggested in the followingsection.

Conclusions and RecommendationsThe theme for the 1973 AIR Forum is "tomorrow's

imperatives today." One of the subthemes concerns"educational program imperatives," which includesprogram evaluation and research on student and collegeoutputs (college success). An imperative for tomorrow inthis area, that has been of more and more concern the lastseveral years. is the need to provide clear-cut evidence togoverning boards, legislatures, and other concerned publicsabout whether an institution is accomplishing all of itsobjectives. Another imperative is that local research not bedone in isolation but that it contribute to overall knowledgeabout students, educational programs and methods, andinstitutions. The previous section outlines some importantreasons why we have not been able to meet such needs atthe present time.

As the decade progresses, and inflation and taxescontinue to rise, the pressures on institutional researchersto do a better job of program evaluation will probably reachunprecedented levels. They will be forced to use moreappropriate evaluation methods and to cooperate more, andthey in turn will put pressure on those in measurement andbasic research to develop more concrete and usefulmeast es of program outputs for institutional use.

What are some possible solutions to themethodological problems outlined? Three possible

Oscar T. Leaning

solutions, which are not mutually exclusive but which areall probably quite controversial, are proposed in thissection.

One possibility concerning the methodologicalproblem mentioned in the preceding section is forgraduate schools to provide a better background in researchand evaluation design and methodology. Such a longtermdevelopment should conceivably help the situation.However, 3 report by Novick (1972), although it wouldappear to encourage better training in statistical methodsand research design, questions whether adequate trainingin statistical methods will solve the problems outlinedunless the skills are used often enough for the researcher tomaintain them at a high level of efficiency. Novick's reportfurther suggests that computer systems may be developedwhich can supplement and possibly even supplant some ofthis desired training. as indicated in the following quoteabout a prototype computer-assisted data analysis system:

(Even those with) substantial training in,understanding of, and competence in statisticalmethods ... are unlikely to exercise those skillsoften enough to maintain them at a high level ofproficiency. These investigators can use and aretypically receptive to guidance in their statisticalwork provided they remain in control of their ownanalyses. Investigators with less statistical skillswill benefit from even more directive guidancethrough the maze of detail required of goodstatistical practice. For all investigators, the tediumof computation or alternatively, the maintenanceof esoteric computer expertise, is a regrettablehindrance to their function of extracting meaningfrom data.

The system described here is an interactivecomputer-based system for assisting investigatorson a step-by-step basis in the use of a particularanatystic tool Bayesian analysts using the twoparameter normal model. The example is meant tobe suggestive of the kinds of computer-assisted da-ta analysis programs that can be developed for useby scientific investigators. Programs such as thesecan also be used in the classroom and laboratoryfor teaching purposes. but beyond this they can beused by the practicing scientist in his dayto-daywork ... An important feature of this program isthat it interacts with the investigator in the Englishlanguage. The investigator need not be familiarwith computer languages or with the internalworkings of the computer. He need only learn howto signin and sign-off the terminal and to makesimple alphabetic and numeric responses. (pp. 1-2)

It would seem probable that such "computer assisteddata analysis system" capabilities could be inexpensively

37

BETTER PROGRAM EVALUATION

distributed to the various campuses through the use of ter-minals from a central computing facility. In addition toguiding the selection of samples, the selection of properstatistical analyses, and the actual computationalprocedures, the system could conceivably be programmedto assist the researcher to avoid pitfalls in the interpretationof his study results. Also, as new and improved statisticaltechniques (e.g., better ways to control for complex inputs)are developed, these could be incorporated into the systemwithout having to worry about their misuse. New technicaldevelopments like this, and computer systems beingdeveloped by agencies such as the Western Interstate Com-mission for Higher Education (WICHE), could be invalu-able aids in meeting the methodological imperatives forresearch in the future.

Another possible part of the solution may be for con-sortia among researchers and among colleges to becomecommonplace, and in each case it would be essential thatthere be good coordination so that the small parts of anystudy would facilitate one another in a planned way. Manyconsortia in the past have failed or been largely ineffectivebecause of poor leadership and coordination, a lack of realcommitment on the part of participants, or related problemssuch as a lack of follow-through and cooperation by participants. Yet there have been some consortia that have beenquite effective and successful, e.g., the Central StatesColleges and Universities (CSCU) Consortium which hasbeen active since 1957. Hopefully, through careful planning(including pilot projects to try out the concept), properleadership, and emulation of procedures used in previouslysuccessful consortia, the pitfalls of many people workingclosely together from widespread localities could be sur-mounted.

Effective consortia could allow expert design con-sultation to take place as a routine matter. Coordination onpredictors, methods, and criteria could allow all of the littlestudies that make up a consortium to add up to a significantwhole that could indicate how colleges of different typesand in different regions of the country differ, They wouldalso allow many of those small institutional researchstudies that are now considered too insignificant to bepublished by a research journal to add to the basic

published educational research literature after meetinglocal institutional research needs. In addition, they couldgive people who are outside of a central I.R. office, whohave access to data but who do not have enough confidencein their tz.search abilities to do research on their own, theopportunity to take part in significant research endeavors.Futhermore, the widespread development of such consortia, each consorium study attacking a specific priorityproblem that has been agreed on by the coordinatingagency or person and the participants, might do away withmuch of the emphasis on "faculty research quantity ratherthan quality." Quality would be built into the system.

Individual initiative within the project and indi-vidual supplemental evaluation outside the project wouldbe encouraged, Certainly you would not want to interferewith the initiative and creativity of the individual resear-cher or evaluator. Participation in such a cooperativeresearch and evaluation effort could also provide excellentlearning experiences for graduate students.

Who would initiate the idea for a particular con-sortium study? Presumably people with particular programneeds who would like to cooperate with others. Profes-sional associations, such as AIR. or other agencies could actas clearinghouses and publicize throughout the country thefact that particular projects had been proposed to study par-ticular problem on college campuses. A regional setupmight be even more desirable.

Whether widespread applications of the possiblepartial solutions proposed here are realistic or not, onlytime will tell, They are presented here primarily tostimulate discussion and thought. There may be otherpossible solutions that hold more promise.

Whatever direction we go remains to be determined,but it seems safe to say that the educational-programresearch imperatives of tomorrow call for new researchpriorities and a new research outlook. Reappraisal andredefinition must take place; new and specific criteria andpredictors must be explored; methodological andphilosophical pitfalls of the past must be avoided; and stepsmust be taken to maximize the practical impact (on thecollege, on students, on college faculty and staff, on society,etc.) of such research,

Domino. C, Interactive effects of achievement orientation and teaching style on academic achievement. ACT Research Report No. 39.Iowa City. Iowa: The American College Testing Program. 1970.

Dunnetle. M.D. Fads. fashions and folderol in psychology. American Psychologist, 1966. 21. 343-352.

Feldman. K.A. Research strategies in studying college impact. ACT Research Report No. 34. Iowa City, Iowa: The American CollegeTesting Program. IWO.

Harris. C.W. (Ed.) Problems in measuring change. Madison, Wisc.: University of Wisconsin Press. 1963.

Novick, M.R. Bayesian considciations in education information systems. ACT Research Report No. 38. Iowa City. Iowa: The AmericanCollege Testing Program. t970.

38

A DETERMINISTIC MODEL FOR INVESTIGATING THE FEASIBILITY OF YEAR ROUND OPERATION

There have been numerous public statenents in thepast several years by business, academic, and politicalleaders favouring the all-year operation of universities andcolleges. In most cases, the suggestion to switch to yearround operation is based on the argument that there is aneed for a more complete and efficient utilization of educa-tional facilities and that the biggest single factor contri-buting to the low overall university utilization is thesummer shut-down of undergraduate instruction. Otherarguments suggest that an all-year university operationwould help students complete their education and becomefinancially self-sufficent at an earner age. while at the sametime removing from the economy the need to provide alarge number of jobs to students during the summermonths.

Mr. A.C. Scrivener, President of Bell TelephoneCompany of Canada, recommended in June 1970 to theCanadian Chamber of Commerce that they give leadershipto a review of university operations. In particular Mr.Scrivener proposed that universities operate two six-monthterms each year and that work-study programs beexpanded. In this way. he felt that better utilization of theuniversity facilities would be achieved and that studentscould more easily earn their way through university byalternately working and studying in these six-month terms.

The Honourable William Davis. then Minis ofEducation for Ontario, spoke in Thunder Bay in Febivary1971 indicating that he looked favourably on proposals forlengthening the university year to permit honours studentsto get their degrees in three instead of four years. TheGlobe and Mail, in their editorial of February 3, 1971,commented favourably on Mr. Davis's remarks. Theycontinued by saying that Ontario's grants to universities forboth operating and construction had increased to $495million that year from $155 million in 1965-66 and that "wecan't stand another five years of high budget increases."

The Globe and Mail continued their support for the"12 -month university year" in a later editorial (written afterMr. Davis became Premier) pointing out that PrimeMinister Trudeau and the Secretary of Stale, GerardPelletier, appeared to favour extension of the universityyear. Mr. Trudeau said: "If we didn't have this type ofschool year we have now which is based on an agricultural

C.J. Christie and D.J. Hipgravo, Woods, Cordon & Co.

society where people attended universities and schoolsduring the winter months and in the summer they workedon the farm in order to pay for their winter it is certain wewouldn't have this bulge of young people looking for workin the summer, and to try to meet that we are trying toconvince the provinces. they they should be more and moreconsidering the full academic year, without this summerbreak."

Mr. F. Lazar, a doctoral candidate in economics atHarvard and regular commentator in the Globe and Mail ontrends in the Canadian economy, wrote in the April 13.1972 issue: "It is generally accepted that the currentstructure of the post-secondary academic year does notfully utilize the existing facilities. Thus the average tax-payer may rightfully inquire whether a plan for a funda-mental restructuring of the school year cannot bedeveloped so that it not only will provide for a moreefficient use of educational facilities (and perhapsmoderate the costs) but atso will improve employmentprospects for students."

These quotations reflect the views, we believe, of alarge number of Canadians who feel that year roundoperation of universities is desirable and that one of themajor benefits of such action is an improvement inefficiency and a corresponding reduction in total costs tosociety. Is this assumption of increased efficiency andreduced total costs in fact correct? Most of the people whomake such statements assume it to be obviously so butseldom have done an in depth analysis of the economicconsequences of year round operation.

In August 1971 the Commission on Post-SecondaryEducation in Ontario awarded a contract to Woods. Gordon& Co. to undertake a study on the "Organization of theAcademic Year." The main objectives of that study were:1. To describe the options in the organization of the

academic year.2. To examine the merit of these choices.3. To estimate the effects of the various choices on

operating and capital costs of post-secondaryinstitutions.

An important feature of this study was to be the useof quantitative modelling to estimate those operating andcapital cost effects.

39

YEAR ROUND OPERATION

Our presentation today will ignore all of the manyother advantages and disadvantages of year roundoperation and will deal with the one issue which isgenerally taken for granted "Is year round operationmow efficient and less costly to society?"

Model FormulationIt became quite clear at the outset of this project that

detailed costs for a university. and in particular costs in a

form which could be related to year round operation, werenot readily available. This ruled out the approach ofattempting to correlate costs with some key variables andselecting the best fits as a basis for the model. The decisionwas made at this stage to base formulation of the model onaggregate cost figures (Statistics Canada, Department ofColleges and Universities of Ontario) and from dataobtained at a Canadian university. The latter was chosenbecause of its experience with the trimester program andbecause detailed cost figures were readily available. Havingformulated this model of a -typical- Ontario university a

40

parametric analysis was undertaken whereby the importantvariables were independently assigned different values andthe effect on total systems cost could be observed. Exhibitshows the key variables that were assigned different valuesor a range of values in the analysis. Other variables areimportant in this analysis but only those which wereallowed to vary are shown. These were chosen becausethey were the variables considered to have economicimportance in going to year round operation.

The model uses total costs over a 20 year period(1971 - 1990) in comparing academic calendar year alter-natives. The method used was to select a parameter set(specific values for total enrolment profile, section sizes,retention rates. plant utilization and net assignable squarefeet per student) and calculate total costs for a number ofyear round alternatives at changing levels of term enrol-ment balance for each year (percent of total enrolmentattending each term) of the time span considered. Thepresent value of these costs are calculated and comparedagainst that for the standard !wo-term program.

Exhibit I

YEAR ROUND OPERATION OF UNIVERSITIESA PARAMETRIC ANALYSIS

NET ASSIGNABLESQ.FT./STUDENT

PLANTUTILIZATION

RETENTIONRATES

ACADEMICCALENDAR

ALTERNATIVE

UNIVERSITYMODEL

TOTAL COST(PRESENT VALUE)

CRITICALSECTION

SIZE

TOTALENROLMENT

PROFILE

ENROLMENTIMBALANCE

LEVEL

SECTION SIZE(CLASS SIZE)

This approach is illustrated in Exhibit II by showinga comparison of the standard two-term program (curve A)against a trimester plan with three levels of term enrolmentbalances (curves B. C and 1)). The upper graph showsmaximum term enrolments for each of these four condi-tions ranging from the standard program with no summerenrolment and a high maximum term enrolment to thetrimester plan with three equal term enrolments and hencea lower value of maximum term enrolment. The lowergraph in this exhibit is a plot of the total annual costs overthe span considered for the four conditions. These costsinclude operating and amortized capital costs, The capitalcosts are only incurred when it is required to purchaseadditional plant and this occurs when the curves as shownin the upper graph reach the existing plant capacity. Hencethe points on the lower graph where the 4:ost curvesincrease in slope coincide with those points on the uppergraph where the maximum term enrolment exceeds theexisting plant capacity and additional outlay is required forplant expansion, This point would occur earlier for thestandard program as shown and later for the trimesterprogram indicating the capital outlays are delayed as thebalanced term enrolment condition is approached, In addi-tion the slopes of curves decrease as the balanced term

Christie and Hipgrave

enrolment condition is approached indicating that thecapital outlays are reduced as the total enrolment is spreadmore equally across three terms. It must be rememberedhere that these are not actual results obtained but rather abasis upon which the model was formulated,

The initial slopes on the curves in the lower graphindicate a base of annual operating costs which increase inaccordance with the increase in the level of total enrolmentover the time span. These annual operating costs initiallyassume that each course is taught in at least three classesand there is therefore no increase in instruction costs ingoing to extended year operation with a trimester program.There is however an increase in the level of the operatingcosts from curve A to D and this is accounted for by theincrease in operating costs other than instructional costswhich are proportional to the increase in the academic yearlength and the extent to which the trimester programachieves a balanced tern. enrolment condition. Exactly howthis is accounted for in the model is explained later in thispaper.

The model calculates the present value of the totalcosts of U, C and I) and relates this to the present value ofthe total cost of A. Assuming that the present value -ofeither B, C or D is less than A,then the indication is that the

Exhibit II

COMPARISON OF ALTERNATIVE ACADEMIC PLANS USINGTHE PRESENT VALUE OF TOTAL SYSTEMS COSTS

MAX TERMENROLMENT

EXISTING PLANT CAPACITY

INCREASING LEVEL W TERM ENROLMENT BALANCE

TOTALANNUAL COST

CIO

YEARS

r,#) STANDARD TWO TERM PROGRAM

,t1 Ce:}4 TRIMESTER PROGRAM WITH VARYING LEVELSOF TERM ENROLMENT BALANCES

YEARS

41

YEAR ROUND OPERATION

trimester program at some level of term enrolment balanceis economically attractive and should be considered as afeasible alternative. However the analysis does not stophere but must go further to convert the results intooperating decision rules. Recall that the assumption wasmade that each course in the standard program was taughtin at least three sections and also that the increase in baseoperating cost in the trimester plan did not include instruc-tional cost increases. This assumption is now relaxed andthe model allows the section sizes to reduce in the trimesterprogram with a corresponding increase in instructionalcosts such that the present values of the two systems aremade equal.

This indicates what section sizes must be achievedin the alternative program to at least make it economicaliyequal to the standard program, The critical value in thisanalysis is therefore section size (class size).

Academic Policy Implications of Model Resultsexperience at universities has indicated that the

PRESENTVALUE

$('000)

145r

140

135

130

42

success of trimester or quarter operation depends onbalancing off the increase in operating costs by the savingsin capital costs with these plans. The high operating costresults primarily from the decision to maintain a full courseoffering in the face of lower term enrolments, particularlyin the summer term. If a university operating a standardprogram has the majority of its courses offered in multiplesections. this increase in operating costs will not be great.However, if the majority of a university's course offeringsare given in one section only, many new sections will haveto be opened and, with the same general level of overallenrolment, the section sizes will be :educed in switching toa quarter or trimester plan. It then becomes of interest todetermine what mean value of section size will make thediscounted cost of the standard program and the alternativeunder consideration equal. These section size values at thebreak-even point are referred to as the critical section sizes.

The critical section size values are directlydependent on the saving in capital costs minus the increasein operating costs for the year round alternative under

Exhibit III

PRESENT VALUE OF TOTAL COSTSVS.

ENROLMENT IMBALANCE

QUARTER SYSTEM(3 ON - 1 OFF - 4 STREAMS)

PARAMETER SET

ACADEMIC LEVEL 1 2 3 4

SECTION SIZES 33 24 21 20RETENTION/ RATES 1.0 .89 .801 .360UTILIZATION IN 1971 90%

STANDARD SEMESTER SYSTEM

($127,937)

.01

BALANCEDTERM

ENROLMENTS

.02 .03 .04 .05 .06

MAD(MEAN ABSOLUTE DEVIATION)

.07 .08 .09

UNBALANCEDTERM

ENROLMENTS

20,000

19,000

18,000

11,000

16,000

45,000

14,000ANNUAL

COST13,000

12,000

11,000

10,000

9,000

8,000

Exhibit ivTOTAL ANNUAL COSTS

VS.

YEARS

PARAMETER SET

ACADEMIC LEVEL 1 2 3 4

SECTION SIZES 33 24 21 20RETENTION RATES 1.0 .89 .801 .360UTILIZATION IN 1971 90%BALANCED TERM ENROLMENTS

Christie and Hipgrave

6,045

STANDARD SEMESTER SYSTEM---- TRIMESTER (2 ON - 1 OFF - 3 STREAMS)-.- QUARTER (3 ON - 1 OFF - 4 STREAMS)

PRESENT VALUEOF ANNUAL COSTS

@ 7t2%

$127,931$123,282$131,787

1974 72 73 74 75 76 77 78 79 80 81

YEARS

'NORMALIZED ANNUAL COSTS PER NUMBEROF FRESHMEN STARTING IN 1971

consideration. This is because the difference is convertedInto a section size reduction to the critical value if the resultis positive and an increase in section size if the result isnegative.

If the difference is exactly zero, then the criticalsection sizes are the section sizes present in the standardprogram and the indication is that there is a fair trade-offbetween the savings in capital costs and increase inoperating costs. Hence, nothing is to be gainedeconomically in switching to year around operation,

The policy implications of this analysis are that theuniversity must regulate its course offerings for thetrimester or quarter plan so that the resulting mean sectionsizes are equal to, or greater than, the critical values.Course offerings for the year round alternative which resultin section sizes less than the critical values indicateuneconomical operation.

In predicting whether an institution could maintainmean section sizes at the critical values if it switched to ayear round alternative. the academic policy-makers wouldhave to determine the number of repeated sections for each

83 84 85 86 81 88

I

89 90

course offering under its present system. It could then bepossible to determine the extent of adjustments in courseofferings necessary to maintain the critical section sizesunder the new plan. For example, if an institution wereoperating under a standard two-term plan and each coursewere offered three times (in three sections) then switchingto a trimester plan would not result in a decrease in sectionsizes and no increase in instruction costs would beincurred.

If each course in the standard program were offeredin a single section and it was decided to switch to trimesteroperation, then instructional costs would triple with acomplementary reduction in mean section sizes to one-third their original-level, If these section size values werebelow the critical values calculated in the model. the onlychoice left would be to drop courses to increase sectionsizes to at least the critical values.

Summary of Model ResultsMany computer runs of the model were made to

show the effects on costs and critical section sizes of

43

YEAR ROUND OPERATION

changes in the key input variables. The approach was tochange the value of one particular variable such as sectionsizes and compare the results of the run for the year roundalternative to the results of the run used as a benchmark.(i.e. the standard semester system). As mentioned earlierchanges were made in total enrolment, term enrolmentbalance, section sizes, retention rates, plant utilization andnet assignable sq. ft. per student. These changes were madeone variable at a time so that the effect on total costs andcritical section sizes to a change in this variable could beisolated.

The following important conclusions are based onthe model results.a) The quarter system with the same freshman enrolment

pattern as all other programs is economically unattractiveunder all operating conditions that were examined in themodel, and should be ruled out as a feasible alternative,Exhibit III shows that the present value of total costs forthe quarter system is greater than the present value oftotal costs for the standard set-nester system for all values

$('000)

130 r

125

44

of term enrolment imbalance levels, This was found tobe true for all parameter sets examined in the model. 'I'osee how this looks on an annual cost basis Exhibit IVshows a plot of total annual cost for the years from 1971to 1990 for the balanced term enrolment condition, Thetotal annual cost curve for the Quarter program nevercrosses the cost curve for the standard semester pro-gram over the years considered.

b) The model results indicate thet there may be someecono is advantage to be gained in switching to a tri-most rogram. As Exhibit V indicates the present valueof the trimester program is less than that for the standard2 term program particularly at the balanced termenrolment condition ($123,182 vs. $127,937). This allowsa maximum reduction in :Average year round section sizesof 6,1% for the trimester program at which point the netpresent values of the two systems are equal. Exhibitshows that the total annual cost curve for the trimesterprogram crosses over the cost curve for the standardprogram in 1975 and remains at a lower value for the

Exhibit V

PRESENT VALUE OF TOTAL COSTSvs. ENROLMENT IMBALANCE

STANDARD SEMESTER SYSTEM

($127,937)

QUARTER SYSTEM(3 ON - 1-OFF - 4 STREAMS)

PARAMETER SET

ACADEMIC LEVEL 1 2 3 4

SECTION SIZES 33 24 21 20RETENTION RATES 1.0 .89 .801 .3E0UTILIZATION IN 1971 90%

.02 .04 .06

BALANCEDTERM

ENROLMENTS

.08 .10 .12 .14 .16

MAD(MEAN ABSOLUTE DEVIATION)

UNBALANCEDTERM

ENROLMENTS

26,000

24,000

22,000

20,000

18,000

16,000

ANNUAL 14,000COSTS

S 12,000

10,000

8,000

6,000

4,000

2,000

Exhibit VI

ANNUAL COSTSvs.

YEARSSTANDARD 2 TERM PROGRAM

PARAMETER SET

ACADEMIC LEVEE 1 2

SECTION SIZES 33 24RETENTION RATES 1.0 .89UTILIZATION IN 1971 90%

3 4

21 20.801 .360

Christie and Hipgrave

+PV (END Of1990) Of ALLOUTSTANDINGMORTGAGES= 14,516

----- TOTAL ANNUAL COSTSANNUAL OPERATING COSTSMAX. TERM ENROLMENTANNUAL CAPITAL COSTS

- - - - - -- - - - -----

3

2

1971 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

YEARS

'NORMALIZED ANNUAL COSTS PER NUMBEROF FRESHMEN STARTING IN 1971

remainer of the time span considered.c) Amortized capital costs make up a small portion of

total annual costs in the university system as determinedin the model. This is caused by a levelling off of studentenrolments over the twenty years simulated and thisfactor has an important effect on the economics of yearround operation. Exhibit VI shows capital cost and totalcost plotted against years for the standard two termprogram for 1971 utilization factor of 90 %. Also plottedon this graph are annual operating costs and maximumterm enrolments. As explained previously, capital costsare a function of the 1971 utilization factor and themaximum term enrolments and for this case account forapproximately 12% of the total annual costs of thesystem al their highest level. As the 1971 utilization rateis reduced, it is reasonable to expect that capital costs arereduced and therefore 12% is the maximum percentageof total costs that is reached.

Since the economic advantage in switching to yearround operation is based on some saving in capital costs

ANNUALENROLMENT

over an increase in operating costs, the absolute value ofcapital costs is an important factor. With the protectedstabilization in university enrolments, capital expansionis no longer required after a certain time (approximately1981 in the model) and therefore the economic advantageof year round operation would be decreased.

d) There is a fair trade-off between savings in capitalcosts and increases in operating costs as the enrolmentimbalance levels in the year round programs are varied.Consequently, the level of imbalance of term enrolmentsfor the year round alternatives considered has little effecton the present value of total costs and critical sectionsizes.

Exhibit VII shows that year round critical section sizesare a flat function of MAD (mean absolute deviation) orterm enrolment level of Imbalance. This points to the factthat as the term enrolments approach the balanced condi-tion. the increased saving in capital costs in changing to ayear round program is offset by a corresponding increasein operating costs.

45

YEAR ROUND OPERATION

PERCENTREDUCTION

Exhibit VII

YEAR ROUND SECTION SIZE REDUCTIONSvs.

LEVEL OF IMBALANCE (MAD)TRIMESTER PROGRAM 2 ON - 1 OFF - 3 STREAMS

PARAMETER SET

ACADEMIC LEVEL 1 2 3 4

SECTION SIZES 33 24 21 20RETENTION RATE'. 1.0 .89 .801 .360UTILIZATION IN 1971 90%

0 02

BALANCEDTERM

ENROLMENTS

e) There appears to be less economic incentive, it any, toswitch to year round operation with low values of the1971 utilization factor.

Exhibit VII shows the effect of assuming variousvalues of utilization rate on the present value of totalcosts. Totai costs fo^ the standard semester program andthe Irimesler program were compared tor the balancedterm enrolment condition for three utilization rafts (60 %.70% . 90% ). The difference in the present value of totalcosts for the trimester alternative minus that for thestandard program becomes smaller and in fact becomesnegative as the utilization rate drops from 90% lo 60% .This is caused by fewer amounts of capital outlay as theutilization rate decreases. in other words, with a1971 utilization rate of 60 %, the increases in studentenrolments through 1990 are served by smaller capital

46

.10 .12 .14 .16 .18 .20

MAD UNBALANCEDTERM

ENROLMENTS

expansion costs than for a starling utilization rate of90% . which reaches its plant capacity at any earlier date.With reduced plant expansion required at the lowerutilization values, the reduced savings in capital costs arcoffset to a larger extent by the same increase in operatingcosts of the year round system. 'fence the allowablereductions in critical section size values are less for eachimbalance level.

1) An increase in the ratio of instruction costs toamortized capital costs. with a reduction in averagesection sizes, indicates that there is less economicincentive to switch to year round operation, When theabsolute values of the average section sizes for theprograms considered were reduced. the present value oftotal costs rose, as did the average value of the criticalsection sizes.

PRESENT VALUEOF

TOTAL COSTS('000)

130

125

120

Christie and Hipgravo

Exhibit VIIIPRESENT VALUE OF TOTAL COSTS

VS.

UTILIZATION RATE

MRAMETER SET

ACADEMIC LEVELS 1 2 3 4

SECTION SIZES 33 24 21 20s RETENTION RATES 1.0 .89 .801 .360

STANDARD SEMESTERSYSTEM

.11['RIMESTER SYSTEM

1

BALANCED TERM ENROLMENTS

SO 60 70 80 90 100

UTILIZATION RATE(PERCENT)

Exhibit IxYEAR ROUND SECTION SIZE REDUCTIONS

VS.

LEVEL OF IMBALANCE (MAD)FOR THREE LEVELS OF SECTION SIZE VALUES

TRIMESTER PROGRAM 2 ON -1 OFF - 3 STREAMS

PARAMETER SIT

ACADEMIC LEVEL

RETENTION RATES

1 2 3 4

1.0 .89 .801 .360UTILIZATION IN 1971 90%

PERCENT2

REDUCTION0

O .02

BALANCEDTERM

ENROLMENTS

06 .10 .12

MAD

SEC TION SIZES ATACADEMIC LEVEE 1 2 3 4

33 24 22 21

24 20 18 17

25 20 IS

.14 .16 .18 .20

UNBALANCEDTERM

ENROLMENTS

47

YEAR ROUND OPERATION

When the absolute value of the average section wasreduced by approximately 15% , there was a corres-ponding increase in the present value of total costs.Exhibit IX shows a plot of percent reduction in sectionsizes versus level of imbalance for three absolute valuesof average section sizes. The percent allowable reductionin section sizes falls as the absolute values of section sizedecline.

This occurs because as the absolute value of sectionsizes decreases, instructional costs and thereforeoperating costs increase. Capital costs remain constantand are a smaller portion of total costs.

g) A reduction in student retention'rates caused a slightlyless than proportionate reduction in present value of total

costs and had a negligible effect on the critical sectionsize values.

The retention rates were reduced by approximately17% and another computer rrn was made to observe theeffect on output variables. While the present value oftotal costs was reduced by approximately 12%, thereappeared to be no change in the values of section sizereductions. This effect is shown in Exhibit X.

h) A reduction in projected student enrolments caused aproportionate reduction in the present value of total costsfor all programs, and did not significantly affect theranking of the alternatives by cost.

The projected freshman enrolments in years 1972through 1990 were reduced by 20% to see what the effect

Exhibit X

YEAR ROUND SECTION SIZE REDUCTIONSVS.

LEVEL OF IMBALANCE (MAD)FOR TWO SETS OF RETENTION RATES

TRIMESTER PROGRAM 2 ON -1 OFF - 3 STREAMS

PERCENT2

REDUCTION0

48

-4

-6

-8

-10

PARAMETER SET

ACADEMIC LEVEL 1 2 3 4

SECTION SIZES 33 24 21 20UTILIZATION IN 1971 90%

0 .02

BALANCEDTERM

ENROLMENTS

.04 .06

RETENTION VALUES

1.0 .89 .80 .361.0 .74 .66 .296

.08 .10 .12 ' .14

MAD.16 .18 .20

UNBALANCEDTERM

ENROLMENTS

would be on the output variables. The enrolment in 1971was left the same so that comparisons might be madewith the present values of total costs derived from theprevious runs. The present value of total costs wasreduced by approximately 20% for each of the programsand the reduction was slightly lower for the year roundoperation alternatives than for the standard semester

10

8

6

4

2PERCENT

REDUCTION

2

4

8

10

Christie and Hipgrave

system. This made year round operation slightly lessattractive economically than it had appeared with thehigher enrolment projections. The smaller reductions inthe present value of total costs for the year round alter-natives caused the allowable reductions in critical sectionsize values to be less for each imbalance level. This effectis shown in Exhibit Xl.

Exhibit xiYEAR ROUND SECTION SIZE REDUCTIONS

vs.LEVEL OF IMBALANCE (MAD)

FOR 2 SETS OF FRESHMAN ENROLMENT PROJECTIONSTRIMESTER PROGRAM 2 ON - 1 OFF - 3 STREAMS

PARAMETER SIT

ACADEMIC LEVEL 1 2

SECTION SIZESRETENTION RATESUTILIZATION IN 1971

33 24 21

1.0 .89 .80190%

0 .02 .o4

BALANCEDTERM

ENROLMENTS

.06 .08

i) A reduction in the net assignable square feet perstudent caused a much less than proportionate reductionin the present value of total costs and slightly reduced theeconomic incentive of year round operation.

When the value of net assignable square feet perstudent was reduced from 130 to 105 (19% ), there was aslight reduction in the present value of total costs for allalternatives (1% -2% ). The reduction was slightly less for

4f RESHMAN REGISTRANTS

PER COMMISSION ON POST20 SECONDARY EDUCATION

.360 20% LESS THAN ABOVEFROM 19/2.1990

.10 .12 .14 .16 .18 .20

MAD UNBALANCEDTERM

ENROLMENTS

the year round operation alternatives than for thestandard semester system because the savings in capitalexpenditures achieved by year round operation werereduced while the increase in operating expensesremained the same. The effect was to reduce allowablereductions in critical section size values for eachimbalance level. This effect is shown in Exhibit XII.

49

YEAR ROUND OPERATION

In summary our model shows that the quarter system is anuneconomical alternative to the standard semester system,that only modest savings can be achieved in switching to atrimester system and that for these savings to be realizedcertain optimum conditions must exist. These include:

(i) The average section sizes for all classes must be atleast at the same level as in the standard two term

10

8

6

4

2PERCENT

REDUCTION 0

-2

-4

-6

-10

system. This may be difficult if not impossible to attainwithout a major alteration of course offerings.

(ii) The physical plant must be presently coping with 75%or more of its potential student capacity under thestandard two term plan without a major capital project.

(iii) The total student enrolment must be evenly balancedacross the three terms.

Exhibit Xil

YEAR ROUND SECTION SIZE REDUCTIONSvs.

LEVEL OF IMBALANCE (MAD)FOR 2 VALUES OF NET ASSIGNABLE SQ. FT. PER STUDENT

TRIMESTER PROGRAM 2 ON - 1 OFF - 3 STREAMS

PARAMETER SET

ACADEMIC LEVEL

SECTION SIZESRETENTION RATESUTILIZATION IN 1971

0 .02 .04

BALANCEDTERM

ENROLMENTS

1 2 3

33 24 21

1.0 .89 .80190%

.06 .08

Future Applications of the Computer ModelThe computer model developed in this analysis

could be used to forecast the economics of year roundoperation for specific universities and colleges. To do this.a detailed cost analysis for the institution under consider-ation would have to be made. This would include breakingdown costs into defined categories as used in the model, orother categories particular to that institution. Enrolmentforecasts would have to be made if none existed, and theeffect on these forecasts predicted should the school understudy switch to year round operation. Retention rates wouldhave to be determined, not only across academic levels butacross departments, to determine if there would be

SO

4

20.360

NET ASSIGNABLE SQ. FT. /STUDENT

130IOS

.10

MAD

.12 .14 .16 .18 20

UNBALANCEDTERM

ENROLMENTS

significant changes by department. The model developedfor this study used average retention rates for the provincewhich differed only by academic level. Some measure ofinstructional space utilization would have to be made forthe institution under consideration and this could involve astudy of class-room use if data were not available. Finally.an analysis would have to be made of course offerings andsection sizes. The total number of courses offered as well

the number of multiple sections for each course wouldhave to he determined. Average section sizes by academiclevel would have to be determined for each departmentwithin the school

ACCOUNTABILITY: SUBJECTIVE IMPERATIVE

There are several fuzzy, ambiguous words whichhave recently become fashionable in the political jargon ofeducation. Accountability is one. Accessibility is another. Itwould be inaccurate to call these "motherhood" words.Motherhood has, after all, acquired a range of fairly specificconnotations. On the other hand, accountability does not byitself communicate any message except that whoever issolemnly using it is solemnly using it. He may know whathe means, although he more probably does not.

Our aim here is to discuss several different kinds ofaccountability:, in operational terms, and to explain whywe think it important for institutional research people tobe discriminating in the ways they serve the principle.We are not attacking the principle of accountability. We doinsist, however, that unless defined and limited, account-ability becomes a popular slogan in whose name all can bedemanded and nothing resisted. We are especially con-cerned with two styles often considered valid for all kindsof accountability, namely input I output analysis employingmeasurable outcomes and objective models relying ex-clusively on quantitative data. In practice, both styles turnout to be an expensive scramble to assemble and analysequantitative date in ways and at levels inappropriate to anyuseful application of the principle of accountability. Theargument for what we are against usually goes like this:"Universities spend vast sums of public money. The publicmust be assured of the benefits of this spending. The wayto do this is to collect as much numerical data as possibleon every aspect of university activities so that the results ofthis spending can be analysed and then justified or con-demned."

Such argument is based on the fundamental assump-tion that accountability must show that resources (orinputs) are producing intended outcomes. We shall arguethat the input I output model of accountability is not onlymisleading but dangerous when used as a basis for asses-sing the activities of universities. We shall suggest an alter-native style of accountability in which the processes of uni-versities are judged at each successive level in hierarchy ofaccountability. This style of accountability recognizes thenature of higher education and is, therefore, likely to bemore effective than any other. It is a feature of this stylethat quantitative data and analysis, white important for

Bernard Trotter and Mario Creel, Queen's University

specific purposes, are subordinate to the role of informedobservation. And this means that at each level of account-ability the level of statistical detail should be compatiblewith the ability and opportunity of those judging to observeat first hand and be aware of the extent to which the datafaits to describe what happens.

The thrust of our suggestions is implicit in the title ofthe paper, "Subjective Imperative." We should thereforetook for a moment at the contrasting terms "subjective" andobjective." We take the term "objective" to mean refer-ence to criteria, standards, facts, which are uniformly trueor held to be universally valid, which are external to andunaffected by a particular system of applications. On theother hand, the term "subjective" means reference tocriteria, judgements, interpretations which have to bereferred to human values or value systems; "subjective"does not mean only the personal or the whimsical, nor doesit rule out intelligent argument or appeal to relevant objec-tive circumstances.

With these definitions in mind we turn first to thequestion: why accountability and what is it? Generallyspeaking, there will always be accountability of some sortwhere there is an express or implied delegation of powerand responsibility. Most institutions, public and private,exercise power and responsibility in order to provide ser-vices or perform functions on behalf of groups of people. Ifthey do not perform as they are expected to, the power andresponsibility delegated to them may be withdrawn ormodified by the delegating group. If there is no delegationcertain goals are agreed at least in broad outline.

AssumptionsIn relation to educational institutions, it is useful to

look at accountability as the process through which mutualconfidence between detegator and delegatee is sustainedand renewed and which provides, from time to lime, thebasis on which the terms and purposes of the delegatedauthority may be altered. This process will rely on much in-formation, non-numerical as well as numerical. Since weare particularly interested in the kinds of information usedat each level of the accountability process we have to askourselves first: why does public discussion lean so heavilyon quantitative analysis of university activities? The

51

ACCOUNTABILITY: SUBJECT IMPERATIVE

answer lies in the fact that almost everyone, including mostinstitutional researchers applies to universities the assump-tions of the input / output model of industrial enterprise. Itis easy enough to understand why this should be so. First,we believe so religiously in the market place that evenpublic spending is thought of and tested in terms ofmarketplace standards. Second, the overwhelming bureau-cratic fashion of the past decade has been program budget-ing and cost I benefit analysis and it is only human toembrace fashion.

Very simply, the assumption underlying programbudgeting is that all public spending programs can andshould be organized in the ways pioneered by Alfred Sloanfor the industrial process. Industrial firms have to plan andrespond to: demand for their outputs, external standards ofquality, well understood external regulations, an advancingtechnology which they (with their competitors) control. Inshort, the industrial world operates on more or less directfeedback. Although large numbers of people are involved,their activities can be monitored and controlled in relationto specific goals production outputs and profits.

In other kinds of activities where the human factor islarge and outcomes are less quantifiable, feedback is erraticover unknown periods of time and the actions of the peopleinvolved are likely to be unpredictable and, more import-ant, uncontrollable. This is particularly true when the deli-very of E., service involves direct interaction between pro-vider and receiver (as in welfare) or rests on assumptionsabout the behaviour of provider and receiver (as in war-fare). In such instances it is easy enough to see why pro-gram budgeting fails. More generally, program budgeting inthe public sector fails because it is based on the hubris ofthe social technician who thinks he is literally ascientist.

Against this background, we can now dispose of theindustrial model so far as universities are concerned,choosing for illustrations one major output universitygraduates. Universities can of course plan size and ex-pected output of graduates, not without error, as we arelearning, but without disaster. But after noses of differentshapes and length have been counted, it is not possible toattach numbers to the output of the university. It is not thatthe multitudinous outcomes cannot be measured with pre-cision, They cannot be measured at all in any ways that dlusignificant for the process of accountability. The most ob-vious reasons for this are: first, the outcome of universityexperience (even if it could be measured for each indivi-dual) is spread over too long a period to provide opera-tionally useful feedback; second, the individuals involvedare themselves playing a part in changing the social contextwhich determines what is to be judged. In short, the cost /benefit model won't work because most benefits cannot bemeasured and others cannot be identified (although theymay be imagined). To complete the comparison with the in-

52

dustrial model, standards of quality are internal to the uni-versity, not external, and they are relative and subjectiverather than absolute and objective. Similar arguments canbe made about another major output the research doneat universities.

A growing body of literature in diverse fieldsprovides testimony that cost I benefit or input I output ap-proaches have not worked for universities. So far. Manypeople studying higher education still have an unwaveringbelief that success is around the corner if only sufficienteffort and imagination is applied. Meanwhile, their wor-ship of trio "quantitative and objective" has led them toproxies for the benefits and outputs in terms of the inputs;that is most advocates of quantitative models for account-ability are in reality using input I input construction. Whatharm this might do we shall probe after we have des'ribedthe process of accountability as we see it.

A Hierarchical ModelIt is helpful to look at the process as a simple hier-

archical model. It further simplifies discussion to concen-trate on the primary teaching functions of universities. Tomake things easier still, we have prepared a table whichsets out for each level of accountability specific examplesof quantitative and qualitative information which may berelevant at that level. We are assuming that each level is ac-countable to the level above for making sensible decisionsabout the resources which have been allocated to it to per-form functions which have been implicitly or explicitlydelegated to it. In other words, the accountability involveddoes not mean demonstrating that specified outcomes havebeen achieved but that resources have been used effec-tively in the pursuit of activities which are judged to haveworth and value. If we could agree that these activitiesthemselves are the product of the university much mis-understanding could be avoided. In any case it is the activi-ties and not the results of those activities which must be as-sessed in the accountability process.

We are assuming a university system in which thereis almost complete delegation to the institution of responsi-bility for allocating resources internally and in which fundsare provided in a lump sum by means of some more or lessobjective formula which does not direct internal budgeting.This is the model we know at first hand. While we acknow-ledge that many others exist in which resource allocation isdecided to a greater extent by authorities external to theinstitution, we would argue that it is both more effectiveand efficient to make decisions about the interrelated acti-vities of universities at a level where the intricacies can beobserved at first hand.

In this hierarchy the most important step in ac-countability is that of the university to the government. Ifyou examine the detail here you will see a very largenumber of quantitative items which are developed in

ternalty so that those making allocative decisions do indeedunderstand the intricacies of the interrelated activitieswhich have to be provided for. You also see a long list ofitems under the non-quantitative information heading.These all have to do with preferences of faculty andstudents, advice from professional groups outside the uni-versity, peer judgements of quality {and while quality stan-dards are determined internally external peer judge-ments are an important factor in the determination ofthese). In reaching decisions about the allocation of fundsto activities, judgements must be made by combining theinformed value judgements of a lot of people with theexamination of a lot of data to arrive at an institutional

Trotter and Creel

judgement. This represents feedback to the budget unitsand Individual professors whose accounting at their levelsabout their stewardship of the resources given to them inthe previous years is a key part of the planning, budgeting,evaluating cycle within the institution. The budget formal.izes decisions about the ways that powers and responsi-bilities delegated to budget units and individual professorsare to be modified. Such changes will reflect changingpr'urities about what needs doing as well as changes inMethods aimed at improving overall the effectiveness of theinstructional process. They may also reflect a judgementthat some activities do not meet established internalstandards.

LEVELS OF ACCOUNTABILITY

DECISION KIND OF DECISION QUANTITATIVE INFORMATION QUALITATIVE INFORMATION ACCOUNTABILITY CRITERIA

GOVERNMENT TO MANPOWER NEEDS; CAREER

ELECTORATE OPPORTUNITY: SOC II TPA

PRIORITIES (INCLUDING

1,. ........ aNa. .... ... .. .. QUALITY 111411,

.--T. EESil OF

SPINDINCGOVt. (RECUT'S,.RATIFIED IA

NUmIER Of STUDENTS,ACTUAL AND FORECAST;

FORMAL DOCUMENTS REPORTS ANDPLANS, ADVICE OF 111001 AND

GOVT EKECUTIVT

TO IMMATUREJUR ISOK TION,WIDE MATCH

10 ABOVE C110004:LEGISLATURE DICKIES SOCHI-ECONOmK PICHESMONA I GROUPS; pus LK EQUITY WITHIN Ott SYSTEM.

AND OTHER DISTRIBUTIONS POLITICAL DISCOURSE AND f EEDBAC lc PUBLIC SATISFACTION;ISTIJOINTI ); AUDITED

......

!SPORTS: GtOSAI INDICES,

......... ......

2. METHOD OF COV1. I XIC VIM INTIM/AMIN? SAMPLING Of IIICOMMENOATIONS OF WEER GROUPS; UNIVIRSITIES SHORT RUN: ORDERETDISTRIBUTION COSTS INTERMITTENT INPUT IOM PlOf 1SSIONS; ADVKI TO 510-GOVEINMENT AND

SAMPLING Of FINANCIAL IOM UNIVERSITIES: STUOENT REQUESTS: GOVERNMENT INTERNAE MANACEMS. NT;{INUITS RECEISID IT WORKING DOCUMENTSor UNIV. LONG RUN; DECREE OfSTUDENTS; PIT TO SOCIEtAt

PRIORITIES AND

PUBLIC SATISFACTION:

1 AUOCAtION UNIVVISITY EMICLITIVI ENROSAIENT PLANNING: DEPMIT PRIORITIES ESTABLISHED Sr SOARDWITHIN

UPEIVIRSITIES

RATIFIED IT BOARD

IADVISIDIW SENATEMENTAL AND StRVICE IUDGETI,

INDICES FOR VN1TS(eSANDOlt UNA'S:STUDENT AND !ACM'S'VIEWS OF T5ACK-01FS1411 GOODS VT

H IKAMERASI STAJI,STUD I NT. AVIIAGE SIRVICES, PSIPMIATION VS. PART.COURSE WAD AND GRADUATES KIPATION, CHOKE VSSTRUCTUREKSkIPE1V1S10t UTHIZA DON RESEARCH POLK t ADMISSIONS POLICY:OF CINTSAI RESOURCES (es METHODS 505 PlIONIOING MIXESCLASSROOMS. 51111451151 AND FAVOR ANC 'HARKS.GIST MOUT 'OBES Of RESIAICH.

FUNDS. STUDENT AID.

TICHNDIOCICAI CAPAIllITIES. UNIT TO

UNISIRSITY

ACAD1MK STANDARDS,MATCH TO VNISIRS171,

ENROLMENT PLAN;11 =memo rm. am... - ......... ..... ...... .,... ,........ ....... ....... ...... ,..,.., ......StUDINT SATISFACTION;S. ALLOCATION UNIT I XI C WM WITH COURSE CATAttS; STAFFING PIII EVALUATION; INDIVIDUAL MI NJOGIMINT;WITHIN ADVICE ANOVII °STAR& TRADIOf I ANALYSES: PREFERENCES FOR /ONO AND STYLE

BUDGET UNITS CONSENSUS STUDENI PERFORMANCE; OF TEACHING; ANO FOS RESEARCH:ENROLMENT PROIEC ?IONS: SWIM NI CH MANE>: DAPASTMINTAI

.,_,.... ,,.... .......

THADOION & ETHOS:

110FIS5O5 PIER IUDGEMENT;TO UNIT STUDENT SATISFACTION;

ACTIVITYIVITY Of PittSONAL WITHIN TEACHING AND SISIAJICH EVALUATION Of AND IT STUDENT: STAFF SATISFACTION:

INOWIDUA 1 INSTITUTIONAL AND OOHS 1%01110UM BUDGETS COURSE OIKCTIVIS. ANTKIPATED

(11410fESS01) SYSTEM CONSTRAINTS 11100115. ?RAVI'. COP/OUTING,ETC-4 RESEARCH GRANTS

PIER 'LOGE AUNT. OEPARTmINTAI

ERPICTATIONS;

53

ACCOUNTABILITY: SUBJECT IMPERATIVE

Data Collection and Its DangersIn the model of accountability that we propose, very

little of the internally generated quantitative data goes for-ward -s part of the university's accountability to govern-ment. The government should hold the univerisity account-able for an effective internal system of accountability andvery little more. The "more" should of course includeaudited financial statements and enrolment information,actual and forecast so that adjustments can be made in theinterests of a rational jurisdiction-wide system. But, pro-vided that the university is demonstrably discharging its re-sponsibility for effective self-government, higher authori-ties should forego detailed management information onfaculty work-loads, class sizes, contact hours etc. The mostfrequent reason for wanting such information is to secondguess the decisions which have been delegated to the insti-tution. This information is wholly inadequate for the jobbecause the detailed data cannot be interpreted by distantbureaucrats or politicians through the lens of first hand ob-servation. And the second guessing will be based on com-parisons with data from other institutions. Since the datawill at first not be comparable for the kind of universitieswe are discussing, government officials then demand com-parable data. M this point the need to define and limit theprocess of accountability assumes crisis proportions. Aca-demic considerations go out the window in favour ofpleasing the data collectors and the central authorities whohave decided that they have discovered a "right" way to dothings. The trade-off here is very clear: an illusion ofgreater accountability less effective institutions. Thepublic is encouraged in its natural belief that "objective"measures of performance exist, and that by feeding theseback into a resource allocation model something like"perfect combustion" in the system as a whole can beproduced.

The chemistry of perfect combustion, if it were not afantasy, would rely on measurement of what happens tostudents on the output side and what faculty do on theinput side. Because the art of measuring student outcomesis acknowledged to be in its infancy even by those whothink it has great potential value, analysis has tended toconcentrate on faculty activities and to assume that thesecan be measured by time spent in doing them. We have al-ready indicated the indebtedness of this idea to industrialmethods. What happens when it is applied? Clocking thetime spent becomes troublesome because academic activi-ties frequently overlap, such as in "preparation for teach-ing" and "research" or in "graduate supervision" and "re-search," and because they are often not comparable, e.g. a"class hour" and a "committee hour." Another oversimpli-fication is that everyone does the same amount of work; animportant fact about universities is that faculty membersdiffer greatly in their productivity. In any practical applica-tion of the "time spent" measure, these difficulties are

54

glossed over by forcing them into an arbitrary fit. This isthe "input" side of the equation. But the time spent is alsoused as a proxy for the outcome of academic activity, i.e. asa measure of accomplishment. This is the "output" side ofthe equation. One can understand the emphasis on whatprofessors do because they comprise the most visible anda legislative backlashnot simply against the method, butask how the money is being spent and how well the job isbeing done, and to answer both questions by pointing totime spent by professors, because the same information ap-pears on both sides of the cost I benefit equation. So, theassumptions underlying the perfect combustion of re-sources, that quantitative measures tell the story, that theycan be put together to provide analytical reports and objec-tive decisions, lead to statements as compelling as timespent equals time spent."

In terms of accountability, of course, the results canbe damaging as well as fatuous. As soon as authorities inthe university or outside indicate that budget decisions areto be made on the basis of scheduled (measurable) time,miraculously the amount of scheduled time increases. Infact it may be both more effective and 'efficient to reducescheduled time. Indeed this has been happening in manyuniversities with the object of providing more informalcontact betwecn professors and students.

But apart from the possibility of unintended dis-tortions which faculty analysis can produce, there Is alarger danger in this and other methodologies which putundue stress on quantitative indicators of performance. Thetechniques themselves come to be seen as panaceas for allthat is thought to be wrong. This leads to expectations ofthe planning I allocation / accountability cycle which aretotally unrealistic and which will, in the long run, producea legislative backlash not simply against the method, butagainst the institutions who helped to create the expecta-tions. It is simply not good enough to say that quantitativeanalysis is still inadequate, but if we put enough effort intoimproving the techniques then we'll have a truly objectiveand reliable standard. We will not. But we will have wastedenormous amounts of the taxpayer's money in the process.

Proper StudiesThe university's accountability should consist only

of demonstrating that it is seriously concerned with makingthe best use of its resources in its particular circumstances.It may be properly questioned from time to time aboutparticular courses of action but it should not be put in theposition of proving that what it has done is "right" or"best" in some absolute sense. Furthermore, those makingjudgements on behalf of the public should not confuselevels of total expenditure with notions of efficiency andwaste. Deciding on "worth" is not the same thing as decid-ing on "waste." An apparently more expensive choice isnot a more wasteful one if a necessary improvement in

quality results.We will try to put these prescriptions into opera

tional terms. The major element in the accountability of auniversity to a government would be to show that the uni-versity's own internal process of accountability willfunction effectively. How can a university do this? Oneway is to make available to the government, or its repre-sentative buffer agent, internal communications whichshow that decisions are being made and are consistent witha real concern for the effective use of public funds. Mostsuch documents are nowadays in the public domain andare prepared for the purpose of informing the universitycommunity about the actions of Faculty Boards, Senates,Board of Governors. From time to time these documentswill refer to and may even describe fully various quanti-tative analyses of activities at the university which havebeen developed to assist judgement in making managementdecisions. The most useful of these analyses are likely to befairly simple and they are likely to be most convincingWien they make explicit the assumptions upon which theyare based. In the form of accountability which we arerecommending, however, government or buffer agencies re-ceiving documents of this kind will not attempt to use thedata for comparative purposes with other universities. Norindeed should they use the data for any purpose of theirown although it would be perfectly in order for them todraw to the attention of the university any flaws which theydetected within the documentation. There is not time orspace here to go into detail but studies such as that pre-sented by Irma Halfter on"A Ration Index of Fiscal -Aca-demic Effectiveness" is a good sample of the kind ofinternal study which demonstrates the right sort of concernfor the use of resources. At the same time the universityshould show that quantitative analysis is a managementtool and not a substitute for judgement. Responsible pro-cedures for budget making is only part of what needs to bedone to convince outside agencies that the university isholding individual departments, agencies and faculties or

Trotter and Creet

colleges accountable each in turn, The university must alsoshow a capacity for internal self-criticism and change,Techniques for systematic review of objectives andmethods in relation to the use of resources of till kindsincluding students themselves have been widely developedin the literature of higher education in recent years.Governments and buffer agencies will want to be increas-ingly alert to evidence that universities are applying syste-matic methods to analyse the correspondence betweenstated objectives and actual objectives, As a corollaryuniversities should be prepared to show that they havefound effective ways of helping individuals attemptinginnovation to offset the conservatism of their faculties. Theuniversity that is truly accountable will be able to showthat careful husbandry does not depend on stagnation.

Fir dly, institutional research can play an importantpart in the credibility of a university's accounting. In somecases, universities have invested heavily in institutional re-search not because they belie/ed in its practical value, butbecause they thought it would make a good impression ontheir political masters and the pubic ct large and because itwas a good thing in itself, It is becoming clear that uni-versity administrations and institutional researchers, andeven more government departments, need to apply to theirown rapidly expanding empires the rigorous systems analy-sis which they insist should be applied to general institu-tional decisions. The institutional researcher who wouldserve his institution and the public interest best is one whoasks before embarking on each projectwhat question amtrying to answer? Is the answer going to be worth the cost?Can I get an approximate answer from data I already have?What does my data leave out? People who use institutionalresearch must remember that educational planning and ad-ministration and teaching all involve the continual interpre-tation of valuesand they must hold on to a confidentcapacity for judgement. Otherwise the search for numbersin place of judgement may end with losing the ability andthe right to judge.

F.E, Balderston and G.B. Weathersby, "FRS in higher education planning and maiinement" parts I, 11 and lif, Higher Education (aninternational journal of higher education and planning). Arnaterdam: Elsevier, Vol. 1 (1972) Nos 2 and 3 and Vol. 2 (1973) No. 1.

f.R Cameron, "Measuring the Quality of Higher Education," STOA (Journal of the Canadian Society for the Study of Higher Education),Vol 11, No 1, 1972, p. 12

Y. Dror, "Ventures in Policy Sciences," New York. American Elsevier Publishing Co., loc.. 1971.

I. Halfter. "A Ratio Index of Fiscal-Academic Effectiveness," institutions! Research and Institutional Policy Formulation (edited byC.T. Stewart) 11th Annual Forum of AIR 1971.

RT. Hartnett. Accountability in ifigher Education, Princeton. College Entrance Examination Board. 1971.

The Learning Society, Toronto, Ministry of Government Services. 1972.

N. MacKenzie, M. Eraut and H.C. Jones, "Teaching and Learning," Paris, UNESCO and the international Association of Universities,1971

55

ACCOUNTABILITY: SUBJECT IMPERATIVE

John J. Pow) Jr, and Robert D. Lamson, Elements Rotated to the Determination of Costs and Benefits of Graduate Education, TheCouncil of Graduate Schools/Washington D.C., March 1972.

Leonard C. Romney, "Faculty Activity Analysis: Overview and Major Issues," Boulder (Colorado), National Center for HigherEducation Management Systems at Western Interstate Commission for Higher Education, 1971.

J. Rust, "The Allocation and Reallocation of Financial Resources to Universities," in (the 12th AIR forum) Reformation and Reallocationin Higher Education,

B. Trotter, D.L. McQueen and B.L. Hansen, The Ten O'Clock Scholar?" Canadian Association for University Teochers Bulletin,Volume 21, No. 3, pp. 4.10.

56

INSTITUTIONAL ACCOUNTABILITY CONCEPT, DOGMA AND REALITY

Underlying most of the current agitation over ac-countability is a profoundly important struggle for influ-ence over if not outright control of our colleges and univer-sities. That struggle is the reality we are dealing with: areality, however, which is masked from view by muchevasive rhetoric both within and without the academy. Inorder to remove some of the obscurities which are hamper-ing our perception, I shall construct and apply some rudi-mentary conceptual lenses, by the aid of which I hope wemay command a clearer view of certain of the problems be-setting us. In particular, it will be necessary to examine therole of certain dogmas and myths = which are ingre-dients in the situation we confront.

I shall, in part, be sermonizing; and my text is takenfrom Lewis Carroll's Alice in Wonderland. You may recallfrom that story the remarkable powers and propensities ofthe Cheshire-Cat in particular, his ability to disappear,leaving behind only his grin. Bothered by the habits of thisstrange creature, Alice asks him to perform this feat oncemore:

"All _right," said the Cat, and this time itvanished quite slowly, beginning with the end ofthe tail, and ending with the grin, which remainedsome time after the rest of, it had gone.1

Current Demands on the UniversityLet me begin my sermon by reminding ourselves,

very briefly, of the truly astounding demands currentlybeing made upon the contemporary university: in parti-cular, upon the large, complex, highly visible, highly dif-ferentiated "multi - versify" or its even more fearsomecousin the mega-versity," vis the state-wide college anduniversity system.

There is a compact and useful summary of thesedemands in the 1972 Wiche report, with which many of uswill be familiar. 2 As the Report correctly indicates (and itis merely summarizing) we are being admonished to"achieve higher productivity" in our educational enterprise;to "attain greater efficiency;" to end "wasteful and un-necessary duplication:** to get on with the job of **measur-ing outcomes;" to demonstrate the "value added" com-ponent in the educational process; to be more "relevant," todevelop and foster "continuing education" (in addition to

W,M, Sibley, University of Manitoba

our existing programs); to be more "open" and "acces-sible," especially to "disadvantaged minorities:" to "endthe abuses of tenure;" to "eliminate deadwood;" to strivefor "flexibility" and to abandon "rigidity:" and, at all times,to be fully "accountable" to "government," to"society," and to the "public" at large. (By some fortunateomission of dispensation we are, apparently, not requiredalso to be accountable to God and to the Queen.)

"Accountability" as the Chosen Instrument of Persuasion1 shall not spend any time reviewing in detail this set

ol ten or more dernaltd3. 1Vc are a!! to familiar with them.Rather, 1 shall fasten on the term "accountability"which, as Howard Bowen has aptly said. is now function-ing as the chosen instrument of persuasion."3 It may benoted that this term does not even appear as a topic in theEducation Index prior to 197(14 Yel it now figures as aprominent theme in a veritable flood of rhetoric: news-paper editorials and government pronouncementsinvariably invoke it. Prior to 1970, it was used to referessentially to matters of fiduciary obligation. Now it isfunctioning as a shibboleth, or as a war-cry; or, to changeanalogies, as a term of incantation to rid the world (andespecially the academic world) of all sorts of presumedmonsters and devils.

Like other currently popular terms (such as"relevance." "alienation," "identity crisis," "participatoryder ,:icy," and so forth) "accountability" is a kind ofct .:1 we d or trigger word for an extremely complex, subtleand shif 'ng array of phenomena, of ideas, and above all ofattitudinal stances or values. Depending upon their usersand their contexts, words of this type shift in elusivefashion, both in their denotation and their connotation. Asheavily value-laden terms, they are most convenient instru-ments of indictment and of attack. Declare yourselves"alienated;" label Classics as "irrelevant;" accuse yourdepartment chairman of being "undemocratic" and"authoritarian" and you can nowadays put in train a setof emotions and attitudes which are most useful as enginesof subversion and disruption of whatever "establishment"or ancien regime you wish to pull down. No impedimentstands in your way; who nowadays will be so crass and ill-mannered as to hold you accountable for the meaning and

57

INSTITUTIONAL ACCOUNTABILITY

use of your words? especially if you choose to talk of"accountability." It is indeed a "chosen instrument ofpersuasion." More bluntly: it is being used as a club.

Let me avoid a misunderstanding here. 1 do not wishto argue that none of the demands being made upon us islegitimate. Nor do 1 wish to suggest that the attacks beingmade upon us are wholly without reason and foundation.We should be concerned, however, about the weight of allof these demands, when summed, upon our institutions ofhigher education. It is calamitous for anyone to require theuniversity, or any institution, to perform tasks it simplycannot do unless his aim is to distort and break it.5 Whatis also of concern is that in this current list of demandsthere is no reference to the tasks which the university isuniquely fitted to perform, has (despite its faults) histori-cally performed, and which no other institution does or canperform. Undiscriminating attack may not reform the uni-versity; it is more likely merely to deform it.

Reality I: The Malaise of the Contemporary UniversityBut is the typical multi-versity in good position to

respond to these demands as it should? Is it capable, inother words, of resisting tho§c demands which should beresisted, and satisfying those demands which arelegitimate? Is it capable of responding in a coherent,systematic and rational manner? No one familiar with ourpresent condition could give a confident reply.

1 do not have thP time to dwell very long in paintinga picture of the profound mat -'se affecting so many of ourinstitutions of higher educations It is obvious to all, andnow a commonplace of many reflections, that over the lastten years or so we have undergone most profound changes.For example, participation in something called "universitygovernance" has altered dramatically in the past decade. Inmany jurisdictions academic stall, .tudents, and occasion-ally members of the support staff now sit on the highestboards or governing councils. Decision-making by"consensus," by "participation," instead of by authoritativefiat is the order of the day.' Yet everyone is disenchantedwith this grand experiment in democratic governance. TheNew jerusalem has not materialized. Everyone goesfruitlessly about trying to find the real locus of power:surely it must exist somewhere? Our students and themajority of our younger faculty are convinced that they donot have an effective share in power, despite theirenfranchisement: and both are quite correct in thisassessment. "Button, button, who's got the button?" theyask in frustratio .1. They are naively convinced thatsomething called the administration" must have thebutton of power. Here they are quite wrong: there is nobutton.

Power has been dispersed everywhere andtherefore is effectively nowhere. "The problem is notconcentration of power but its dispersion. Inertia,

58

incapacity to respond, is the inevitable outcome Nolonger do we have, in any genuine sense, a system; we haveonly what one of my academic colleagues aptly calls "this.crazily juddering heap."8 The old "community" hasdissolved into adversary relationships: we have moved (toquote Nisbet) from legitimacy to litigation.10

A useful way of highlighting this absence ofcoherence or of system is to consider the quite incrediblearray of models through which the various protagonists inthe struggle for ownership and mastery of the universitywant to see (and persuade others to see) our institutions ofhigher education. At limes, some remnants of our founda-tion model come into play: the model of the guild ofscholars, a model, as Nisbet's magisterial analysis shows,rooted in an essentially mediaeval context. This is themodel which sets in play much of the rhetoric about "thecommunity of scholars," "university autonomy." the"sanctity of tenure." and like phrases.

But other models are not wanting. There is theconsumer sovereignty model; the public utility model; thesocial service station model; the model of the university as"therapeutic community," functioning as a sort of"psychiatric day hospital;" the corporate model; and avariety of political models. These diverse models (orrelevant portions of them) play like magnets over theelements of the whole, aligning them now in this array,now in that, but with no magnet possessing sufficientintrinsic power to override the attractions of the competingmagnets so as to impose and maintain any coherent.permanent recognizable pattern or system. In the accurateusage of mathematics, the university of today appears to beapproaching "the degenerate case;" i.e., to be in danger oflosing those characteristics proper to its genus or nature.Power to respond in a rational and coherent way to theforces acting upon us in the name of "accountability" isbecoming small or nonexistent.

Conceptual and Linguistic Pollution and its EffectsThe university is by no means wholly responsible

for the condition in which it now finds itself. It has beencaught, willy-nilly, in a vortex of social, economic andpolitical forces which are in large part not of its ownmaking. Yet we are not wholly guiltless either, for theprocess of dissolution has been aided and abetted fromwithin the academy itself. We have permitted ourintellectual and linguistic coinage to become debased. Or.to borrow terms from the realm of ecology, just as ourphysical environment has suffered pollution from ourabuse of it, so has our intellectual environment suffered acorresponding decay and pollution. And that is somethingwhich should not have happened in the academy, of allplaces.

The fact is that we see our environment throughwords and concepts and we act on it in terms of what

we see. When some one coins a striking word, we seethings differently, and we then respond differently. Once aterm like "identity crisis" becomes established, adolescentsstart having "identity crises" by the millions. Allowwords and concepts to become muddled and slovenly, sothat the lenses through which we see become clouded anddirty, and we shall soon begin to stumble and to actfoolishly.

Let me illustrate by reference to the context ofuniversity affairs. I am rather confident that a generationago we did not go around talking of the problem of "univer-sity governance." Very unfortunately, that phrase was notsmothered in its cradle: it was permitted to grow and loacquire undue influence over our vocabulary. By insensibledegrees, we then began to speak of "the administration,"using the term to embrace such offices as those of presidentor principal, dean and department head. We thereby (quiteunconsciously slipped into two serious errors. In the firstplace, we began to obliterate the former clear distinctionbetween (say) heads and deans on the one hand andcomptrollers and similar functionaries on the other.Secondly, under the influence of bad analogy, we began tospeak and to think of all these "administrators" as clerks or"civil servants" a change not muted in significance byinsisting on the (relatively trivial) distinction that somecivil servants are more senior than others. Logic (of a sort)now carries us on to the denouement, wherein we find"academics" asserting that these mere "administrators" (arethey not only civil servants?) should be excluded from anyreal influence of senates or boards of trustees. Surely this isAlice in Blunder land? Somewhere along the line, we havequite forgotten that any institution requires leaders. Indeed,words such as "leadership" and "authority" have nowacquired such an odour as almost to be unmentionable inpolite academic society: one must apologize for using them.The "anti-leadership vaccine" of which John Gardner wroteso perceptively seems to have done its work only too wel1.12

am not here engaging in any fruitless exercise innostalgia. (As the wry graffito has it, nostalgia is not what itused to be.) I am simply recalling that, a generation ago oreven less, the university was embodied in its heads, itsdeans, its principal or president. Those office-holdersspoke for the university; they were (or were taken to be)exemplars in their own persons of what the university wasfor. They were the moulders and makers of the university;they were responsible for its nurture and its growth.

In a word, they were the authorities. In a preciselanguage such as Latin, they would have been calledauctors. Auctor is related to auctoritas, "authority:" andboth are related to the verb augeo. "to cause to grow orincrease." In our slovenly modern manner a ;peaking andthinking, we have lost both the word and the concept. Inthe enstpg development, we have, quite naturally, lost ourauthorities, our "makers." Of "un-makers," of destroyers,

W.M. Sibley

we clearly have a plenitude. Authentic decision-makingbecomes more and more difficult, Here and elsewhere,within the university and without, the paramount problemis the unmaking of indecision.13

"Accountability" Some Conceptual Clean-UpIn destroying authority, we have simultaneously

destroyed the necessary conditions of accountability. Youcannot talk meaningfully about accountability without talk-ing about authority. The notion ci accountability entails thenotion of authority, of a structuring of "offices" withinsome juridical or institutional framework. Yet many ofthose who insist on more of the formr are doing their bestlo ensure that there is steadily less of the latter.

What, in standard English, is the meaning of "ac-countability?" Consider some office-holder, X; for example,a faculty dean. If X is to be held accountable for thedischarge of some duty or trust, then there must be somesecond person (or body) Y, (e.g the president) who can"call X to account:" that is, who can demand that X justifywhat he has done or has failed to do. if X refuses to givesuch an account, or if Y finds his account wanting in somerespect, then Y is empowered to inflict upon X an ap-propriate penally or "sanction," or to request some otherauthority to do so. In plain English: the dean can be fired.Remove such notions as "sanctions" and "authority" andyou remove all meaningful talk about accountability. Wemay put this in another way. Impose accountability whileal the same time destroying authority, and (if you are sominded) you have al your disposal an admirable engine fordestroying the bonds which hold an institution together,which make it a system.24

To give force to these comments, let me contrast twovery different contexts for decision-making. Let us firstimagine some office-holder, X, operating in a well-structured system, who makes (in responsible fashion) adecision likely to create some painful consequences. Someof the shocks flowing from that decision will inevitablyrecoil on him. The action is reflexive: for the action, there isa re-action, as the first law of Newton assures us therealways must be. If he has made a mistake, he is not onlyblameworthy; he will get the blame or receive the penaltyin his own person. When blame can be fixed in this way,accountability is present. X's action and role are systemconserving.

Suppose, secondly, a committee (not an individualoffice-holder) is called upon to make the decision. Im-mediately we are in difficulties, Unless we blandly reifythe committee, there are notable difficulties even inspeaking of an action and therefore a decision at all. A votecan of course be taken; it may even be unanimous; butthere may in practice be as many reasons, all different, asthere are members of the committee. There may becomplete formal consensus as to the motion, but corni:Iete

59

INSTITUTIONAL ACCOUNTABILITY

dissensus as to the rationale. Accountability immediatelybegins to come unstuck, because there is no authoriativeway even of providing an account of the action.

A second difficulty is that the necessary condition ofa sanction is removed. In a loose sense the committee canbe "blamed:" but there may well no way, short ofexecuting the whole lot, of bringing a sanction to bear on acollective or on its members. The action is not reflexive:blame can never be fixed upon any individual member of acommittee. Even if it could, what appropriate sanction isavailable? Being fired from a committee. as we all very wellknow, is no sanction at all. Even if sanctions could bedevised, by the time the adverse consequcw.as appear thecommittee's membership is very likely to have changed. Orthe committee may even have ceased to exist. It is about asuseful to blame a committee as to blame the hurricane fordestroying your house. To say that the committee "acted" oris "accountable" is to begin to use words in a degenerate orPickwickian sense. Relying on committees for advice isprudent; relying on them for decision-making is not system-conserving but system-destroying.

The Masked and the Masking GameThe realities of our situation in the face of she

demands of accountability even the legitimate. not thespurious demands are not particularly reassuring tocontemplate. For several reasons we tend to conceal theserealities in fogs of various kinds. How and why do we dothis?

Let me begin with a very simple example one nodoubt recognizable to all who have seen academe in action.The Faculty of Arts at Puddleberry University is engaged inthe throes of its quinquennial debate about the curriculumfor the Arts degree. The Department of Outer MongolianLanguage and Literature does not have quite the stranglehold on the curriculum which it enjoyed at the time of Pud-dleberry's founding. Some bold spirits in the Department ofIdeology have therefore advanced the proposition that OM200 should henceforth not be required for the degree. Thatmotion is the formal agenda for the meeting of the FacultyCouncil.

Professor Hapless, Head of Outer Mongolian. willhave to repel this thrust of the barbarians against theramparts of high culture. Everyone but a babe in armsknows that Hapless (is he not human, like the rest of us?) isvitally concerned about losing student credit hours (shouldstudents be allowed to vole with their feet): academic andsupport staff entitlements; resources of other kinds. Hisstatus and clout in the Faculty are at stake. The struggle atthe Faculty Council between Hapless and the ideologistsover these matters constitutes "the hidden agenda" a

phrase now a commonplace.Can Hapless address himself to the real or the

hidden agenda? Obviously not. The unfortunate man is

forced to enter upon a debate about the intrinsic merits,high educational values, etc., of OM 200. He must disavowany self-interest in the issue, and advance arguments of apersonally disinterested, moral and rational type, addressedto some shared (or once shared) concept of the mission ofthe Arts Faculty of Puddleberry University. Given theseconstraints, Professor Hapless clearly has a formidableassignment in front of him: and smart betting is that he willlose.

At this point I want to introduce the notion of agame. I shall not depart from the standard dictionarydefinition of a game as "a contest played according to rulesand decided by skill, strength or luck." That is a simpleenough definition. There is. however, an incredible varietyof games, depending upon the kinds of skill, strength orluck at issue and upon the pertinent rules; and full exploration of the concept is a most demanding task." Here I shallhave to avoid complications, noting only that we makeeveryday use of the word in talking about the actions andbehaviour of others. "What's his game?" "What's the nameof the game?" are two obvious examples.

I shall call the "hidden agenda" the hidden or themasked game. in its crudest form, the masked game isconcerned with the simple question; "What is hest for me?"Professor Hapless, however, is forced by a certain set ofrules to play another game, which covers up this somewhatunpleasant, naked reality. Let us call this second game themasking game.

The masking game always has a standard form. Itsdetailed content will vary with the context. The standardform is: "What is best?" or "What is best for . . .?" wherethe blank may be filled by designating any group or anycause transcending patently "selfish" concerns. "What isbest for the students of Puddleberry?" is a move open toProfessor Hapless. What is best for the nation? for theParty? for the Revolution? are other possible fillings for thebasic form. The essential feature is that within the maskinggame, we must always set in play some admixture of themoral and the rational.

Everyone, including Hapless, is aware that he isplaying the same game a fact not concealed from them-selves, their audience, or from Hapless. (We may be confi-dent, for instance, that they will be engaging in much talkabout "relevance" as a mask for their real interests.) Whythen all the pretence? Why does not Hapless simply standup in the Faculty Council and state his teal concerns? Theanswer is that this option is simply not open to him. if hewere to drop the mask, and admit to his real game, no oneelse would play with him. There are severe penalties fordropping out of the masking game. To refuse to play by therules and constraints of that game is either to leave or totake on the whole organization or institution of which he isa member. Almost certainly Hapless cannot afford to optfor either alternative. He is therefore forced to play, with

the big guns of "relevance" trained straight on him.If you think that 1 am going to condemn the masking

game, you are quite wrong. It is only the extreme and de-generate case of the game that I shall condemn. In fact,playing the masking game (as 1 have defined it) is an in-escapable and essential feature of the human situation.Aristotle put it succinctly when he laid down the funda-mental proposition that by nature all men are "political" orsocial beings. We are neither beasts nor gods; we are half-way between. We are not driven merely by brute instinctsfor survival. Nor are we like the gods perfectly moral,rational beings. We are in a middle condition; and we areforced to live within the bonds of some society or asso-ciation. Self-interest must be curbed, muted; disciplined bymoral and rational considerations if that society is toendure. In our discourse with one another we must put onthe mask of morlity and rationality.

Certainly. the masked game influences, perhapsdistorts, the play within the masking game, (Some movesare just of available to Professor Hapless.) But equally, themasking tame moderates the play of narrow self-interest,which is the name of the hidden or masked game. In ahealthy, genuine community. the two games interact. Themask modifies its wearer. Your three year old child doesmore than merely don the Halloween mask of the witch: forthat magic hour she is a witch. The mask or persona modi-fies the person: persona, indeed, originally meant the maskof the actor in the play. The role called for by the maskchanges us, a fact noted long ago by William James in hisPrinciples of Psychology. Assuming a virtue when we haveit not is the necessary first step in practising the virtueand therefore in coming eventually to possess it.

"Valence Bonds" and the Sustaining DogmaThe elements of any viable organization are held to-

gether by certain "valence bonds" by certain values. Themasking game must appeal to those value bonds; as long asit does so it is a necessary and indeed value-reinforcingactivity.

Those values will be found to be embedded in acentral, sustaining doctrine or set of fundamental beliefsheld as valid, and functioning as dominant and unquestion-able for the particular organization. Such a doctrine is adogma. No doubt at some time Puddleberry had a local andcompelling version of the doctrine of the community ofscholars; and that unquestioned doctrine with its associatedvalues and beliefs conferred rights and status upon certaincentral disciplines and their professors. As long as the sus-taining dogma remains strong, the barbarians will probablynot dare to invade the realm of Outer Mongolia. They willindulge only in reconnaissance forays. But once weaknessis scented, they will attack. That is why, sooner or later,Hapless is bound to lose. Life has already fled from the sus-

WM, Sibley

taining central dogma. The case of Hapless is hopeless:avoid him at all costs.

What happens when an organization begins tocrumble? when its core of values starts to dissolve and thevalence bonds to loosen? when community is replaced byadversary relationships and new power alignments? whenan institution loses those qualities proper to its nature, i.e.,when it approaches degeneracy? Then the masking gameitself becomes degenerate. The moral and rational dialogueforced upon us by the rules of the masking game ceases totame and to mute self-interest and partisanship. The sus-taining myth or dogma may eventually come to be used inthe most cynical manner by the contending factions.

Consider, by way of example. the classic descriptionof the university as a "community of scholars." Perhapsthere was a time when in most if not all places that expres-sion denoted a genuine reality. No doubt there are placeswhere it still does. As applied to the typical large, modern,urban multiversity, however, the use of the expression nowmakes one distinctly uncomfortable if taken as a refer-ence to a governing reality. As one acerbic writer puts it,most debates about :'university governance" are "part of afantasy about the existence of 'community.' But large urbancampuses have about as much community as Grand CentralStation:46 The expression in such contexts denotes a merefiction: in current usage of the term, a myth. And its use inthe masking game may become more and more cynical andunprincipled.

For example, over the last few years in many institu-tions the "community of scholars" argument has been used:(i) by the faculty. to wrest power from the "administration"and from the board of governors or trustees; (ii) by faculty,with the administration as opponent and the board as allyor neutral; (iii) by students, with the faculty as opponents;(iv) by students and faculty, with administration and boardas opponents; and so on through still other game combi-nations. Recently, at my own institution (perhaps others?)the local faculty association made an interesting externaluse of the myth a use which some of my academic col-leagues have found somewhat cynical. Applying to the pro-vincial Labour Relations Board for formal certification as aunion. they claimed in their application the president, theacademic vice-presidents, and all deans as members of theunion on the ground that since all were members of a"community of scholars" all had that true community of in-terest into which the Labour Board is bound by statute toinquire. The contradiction is obvious: if we are all membersof a community, what are any of us doing at the Labour Re-lations Board whose role essentially is to organize the,fight?

Reality The Accountability InterfaceThus far I have been talking about the institutional

61

INSTITUTIONAL ACCOUNTABILITY

scene, and noting the conditions which tend to defeat ac-countability within the multiversity. Some brief commentsabout the interface between the university and the govern-ment are now in order. Hero too we find a masking ofreality which is causing serious if not intractable difficulties for those who have to plan and manage ("manage" isnow the operative word) our colleges and universities.

Governments nowadays are fond of posturing champions of "accessibility:- as enemies of "elitism:" as true be-lievers in full "self-realization" through education for allcitizens of the state; as committed adherents of "participa-tory democrac:i "openness of decision-making." and thelike.

But what is the reality? it is a commonplace thatevery good medical school on this continent is turningaway five, six, ten, twenty or more wellqualified appli-cants for every student admitted. We admit only a highlyselected group: not to mince words, an elite. We may fairlysuppose that many of those rejected would have foundgenuine selfrealization in a career in Medicine to saynothing of the fact that they are also badly needed. But askthe government to fund you at a level which will permityou to treble or quadruple your medical faculty intake, andthe talk about universal t,..cessibility is suddenly turned off.II turns out that governments are interested in accessibilityat a price; and that they are unwilling to pay the full priceof universal accessibility. There are hidden or masked constraints on their allegiance to "accessibility," As for opendecision-making, one has yet to hear of the party caucus orthe government cabinet being thrown open to the press.The realities of political competition simply rule that out.Governments, again, insist that they believe in efficiencyand productivity while their civil service rolls expandevery year at astonishing rates. Yet universities are con-sidered fair game for all those demands we noted at theoutset which are embedded in the popular cry for "account-ability.-

I am not here trying to belabour governments: rightnow, most democratic governtnents are struggling with ahost of perhaps unmanageable problems and are in a caseas sad as that of the multiversities. Faith in our institutionsgenerally is badly eroded. I am merely trying to indicatethat cant and hypocrisy are by no means the private pre-serve of academe. Governments also play and of neces-sity thn masking game; and in our own legitimate in-terests and that of our institutions of higher education it ishigh time to start saying so.

ConclusionLet me review and conclude. I have been arguing

that the university is generally in poor condition to respondto the demand for accountability, even whete.that demandmay be legitimate and not merely a club to beat us into sub-

62

mission. For this condition there are many causes: someinner, some outer. One pervasive cause is that the rapidgrowth forced upon us in the decade of the sixties was toomuch for any organization to cope with. As Sir Eric Ashbyhas wisely remarked, it is the experience of a biologist thatlarge mutations are invariably lethal: the same law appliesto social organizations.17

Whatever the causes, much of the substance of theold order has vanished. Let me now advert to Alice inWonderland and modify our text by one word only:

"All right," said the University, and thistime it vanished quite slowly, beginning with theend of the tail, and ending with the grin, which re-mained some time after the rest of it had gone.

Unlike the CheshireCat, however, we cannot reappear.

What, then, is to be done? No short, easy answers todifficult problems can be expected. People who have theanswers, one finds, usually don't have the problems. Thesimple mentality of the Queen of Hearts, for whom thecommand "Off with his head!" was the recipe in all diffi-cult cases, will not suffice.

What we face is a long and difficult task of recon-struction. And one of the first essentials is to start talkingwith much greater candour about our present situation. Aslong as we persist in presenting ourselves to the public asan autonomous community of purely rational, disinterested,high-minded scholars, dedicated only to truth, governingourselves in dispassionate and efficient fashion, we aregoing to be in trouble. The public will naturally (andrightly) wonder why, if that is the case, we cannot orderour affairs in better fashion. As Warren Bennis has noted,all of our institutions including the university arefacing an increasing credibility gap. Lack of candour, hesuggests, fosters "the adversary culture;" and moreover"has a devastating impact on the effectiveness and moralityof our institutions...Is

Some of the outlines of the future are very dim.Others, I think, are tolerably clear. There is a well-knownlaw that as society evolves, its members tend to move froma condition of status to one of contract. No longer can wetalk, as philosophers of a previous generation did, about"my station and its duties". The movement from status tocontract has now caught up, with the university. We are atthe end of an era; and in our reconstruction we shallprobably have to build upon new contractual bases ofwhich faculty unionization, now the focus of so much in-terest. may well be typical. In any event, it is quite certainthat much of what we have known and valued is going topass away, if it is not already gone. With proper leadershipand resolution, however, we should be able to salvagesomething of the central core of our tradition and values.

Only, however, on the condition that we examine

WM. Sibley

ourselves more critically and address ourselves more vigor- living. My addendum is more modest: it is also just tooously to repudiating outworn myth and dogma. It was dangerous.Socrates who said that the unexamined life is not worth

s Alice in Wonderland, Chapter VI.

1 Annual Report, 1972, Western Interstate Commission on Higher Education (P.O. Drawer P, Boulder, Colorado, 1973). pp. 2.3.

3 Bowen, H,R. "Holding Colleges Accountable," Chronicle of Higher Education, Vol. VII, No. 23, (March 12, 1973), p. 28.

4 Mortimer, K,P, "Accountability in Higher Education." ERIC Clearinghouse on Education, Vol. 1, 1972, p. 47.

s See Annan, Noel. "What are Universities For, Anyway?" The Listerer, Vol. 88, No. 2275, (November 2, 1972), pp. 597-600.

The malaise is accompanied by a crisis of confidence. It is only fair and important to note that confidence in other institutions Issimilarly eroded. Over the last five years the erosion has indeed been precipitous, as pointed out by Harold L. Hodgkinson in his address to the1972 Forum of the Association. (See Hodgkinson, H.L., "Open Access A Clue to Reformation and Reallocation:" in Reformation andReallocation in Higher Education, 12th Annual Forum of the Association for Institutional Research, (1972), pp. 10.16.)

For a most illuminating discussion of the possibilities and of the very great theoretical and practical difficulties lying in the way -.of participatory modes of government, see the "curriculum essay" by David K. Hart, "Theories of Government Related to Decentralization andCitizen Participation." in the special issue of Public Administration Review, Vol. XXXII, (October, 1972), pp. 603-621.

Sibley. W.M. "Accountability and the University: Is Whirl to be King?" Stoa, (Journal of the Canadian Society for the Study of HigherEducation). Vol. 2, No. 2, (1972), p. 10.

Carney, T.F. "Towards 1984." Unpublished submission specially commissioned by the BonneauCorry Commission on theRationalization of University Research in Canada (1970), p. 42.

so Nisbet, Robert. The Degrodation of the Academic Dogma, (New York: Basic Books, Inc., 1971), pp. 43-44 and 137-138.

ss See Vann, Carl R. and Singer. Philip. The University as Therapeutic Community." Change. Vol. 4, No. 5, (lune, 1972), pp. 6 and 58.59.

11 Gardner, john. The Antiteade rshlp Vaccine." 1965 Annual Report, Carnegie Corporation of New York,

14 I owe the phrase to Professor T.F. Carney, of the Department of History. University of Manitoba.

11 For further development of the concepts, see the writers article in Stoa, cited above.

is Ludwig Wittgenstein's Philosophical Investigations is a most powerful and influential exploitation of the concept of the game, in thefield of Philosophy. For an example of the influence of Wittgenstein's treatment of games in the field of political science, see the penetratingand impressive study by G.T. Allison. Essence of Decision (Little. Brown and Co., Boston, 1971), particularly at the level of development of histhird "analytic paradigm" in Chapters 5 and ff.

IN Goodfriend, H.J. "The University as a Public Utility," Change, Vol, 5. No. 2 (March, 1973), p. 62.

It Ashby. Sir Eric, as quoted in Change, Vol. 5. No. 1. (February. 1973), pp. 22.25. (The reference is to Ashby's Any Person, Any Study(McCraw-Hill. 1971).

la Bennis, Warren. The Leaning Ivory Tower, (fossey-Bass: San Francisco, Washington. London, 1973). p. 13. (See also Bennis' incisivecomments about the difference between "administration" and "leadership:" and his penetrating remarks about the serious obstacles to genuineleadership in the contemporary university, at pp. 82-84.)

"a, I",

THE INSTITUTIONAL RESEARCH IMPERATIVE AT EMPIRE STATE COLLEGE

A. Paul Bradley, Jr., and Ernest C. Pulpit), Empire State College

Seldom in higher education has a college or a uni-versity received at founding a mandate to conduct large-scale institutional research and to use findings for im-proving institutional functioning, In fact, even at existingcolleges and universities, institutional research is often alow priority Indeed. The new, nontraditional college of artsand science of the State University of New York is a not-able exception, for, at Empire State College, research andevaluation claims equal status to the other key institutionalfunctions. This paper will give an overview of Empire StateCollege, discuss the rationale for institutional research, andoutline the projected research and evaluation program.

Overview of Empire State CollegeEmpire State College is to serve as an alternative to

the traditional institutions of New York State. It representsa " . .statewide commitment to test and experiment withnew, flexible, and individualized modes of learning."' Be-cause of this charge. Empire probably has more unusualfeatures than any other single institution in the UnitedStates or Canada.

Empire is a college without campus. Instead. theinstitution utilizes a statewide network of mechanisms forreaching students. The most prominent mechanism is theregional learning center which generally has a dean, anassociate dean, an assistant dean, and thirteen to sixteenmentors covering a variety of disciplines, though otherstaffing patterns also exist. Currently, there are regionallearning centers in Albany, Rochester. Manhattan. andLing Island (Old Westbury). By 1974.75, the Collegeintends to complete statewide coverage with the addition ofcenters in the lower Hudson area, in western New York. inthe southern tier, and in central New York.

Regional learning centers are the primaryadministrative and academic loci for planning and ful-filling student programs of study. Operating year-round,they have responsibility for developing and extendingvarious educational services throughout their entire regions.In doing this, centers will increasingly find it convenient toestablish satellites at such locations as public and privatecolleges, museums. industrial plants. prisons, community

64

centers. churches, and governmental agencies. Satelliteshelp Empire State College go to the students and will beexpanded or discontinued as student need and demandrequire.

The overall administration of Empire State College islocated at the Coordinating Center in Saratoga Springs.There the President, the Vice Presidents, and their staffssupervise the formulation and administration of college-wide academic, fiscal, personnel and facilities policies inconsultation with the deans and faculty.

The idea of Empire State College requires thatfaculty seek ways to provide study opportunities in re-spOnse to student's individualized purposes, interests, andneeds. Furthermore, the College is to use the variety oflearning resources that exist in the State rather than dupli-cate existing resources. This translates into essentiallythree modes of learning: the mentor mode, the adjunctmode, and the organized program mode. In the mentormode, students come to the regional learning centers orsatellites and meet face-to-face with faculty (calledmentors). Together the mentor and student develop and im-plement an individualized program of study built on aseries of learning contracts. These programs of study mayinvolve elements of the adjunct mode: field studies, tutors,regular courses at other colleges and universities, workstudy opportunities. internships, and the like. It also mayinvolve use of organized programs which are self-studymodules developed by the Learning Resources Faculty, ahigh-powered group of scholars recruited both nationallyand from the regional centers and housed in Saratoga.These organized programs can be used by students whohave mentors and by students who wish to work inde-pendently.

Another specific element of the Empire State Collegeidea is that students be granted significant amounts ofadvanced standing on the basis of prior learning. Thislearning may have come about in formal academic situa-tions or simply through the experiences of a lifetime. Stu-dents document their learning in autobiographical port-folios which are assessed at the regional learning centers

and the Coordinating Center. Currently, a student can earnup to five-sixth's of his degree on the basis of priorlearning.

Rationale for Research and Evaluation at Empire StateCollege

One of the early decisions made at Empire State wasto have a strong internal research and evalurtion program.Such external groups as accreditation commissions and thelarger educational consultant firms could and mayeventually conduct a number of useful studies. However, acommon theme in relevant literature is that internal re-search programs have advantages over external researchprograms in promoting organizational change. As Burnsand Stalker= and Lawrence and Losch3 point out, thestructure of organization should be contingent upon thenature of the organizational environment. For EmpireState's dynamic environment, Peterson4 would draw uponthese studies and recommend an active internal researchoffice to provide systematic and regular input to the de-cision making process. Important to this argument is thenotion of acceptance of results. Before they will promotechange, decision-makers must "internalize"s the need for it.Achieving this can be difficult so Bennis recommends that

. the more profound and anxiety-producing thechange. the more a collaborative and close relationship isrequired."s French and Raven' would agree, noting "refer-ent-power" idcntificatitn with a person's viewpoint be-cause of feeling for and understanding of that person isa likely outgrowth of regular interaction. Thus, as an inte-gral part of and a regular contributor to the decision-makingprocess, the research and evaluation program at EmpireState should play a vital part in the fuhre development ofthe College.

A second reason for having a strong research andevaluation program at ESC is the great interest in the in-complete but evolving plans and programs of Empire StateCollege. Students, faculty. ESC administrators, State Uni-versity administrators, legislators, high school guidancecounselors. and many others across the country are askingimportant questions of the new venture:

What are the characteristics of students at-tending the College?What sorts of activities do students engage induring their learning careers at the College?What are the activities of faculty and adminis-trators in the Coltege, and how do they differfrom those of people at more traditional in-stitutions?What are the outcomes for students and otherswho participate in the program and processesof the College?What can be said about student and faculty

Bradley and Mole

attrition from the College?What are the costs associated with this enter-prise and with its major elements: assess-ment of prior learning, contracts, organizedprograms?

The challenge in dealing with these and other simi-lar research matters is that the "correctness" of the answersdepends upon the perspective of the questioner. Forexample, in reporting on the mentor-student relationship.mentors will want us to emphasize the teaching that goeson, SUNY administrators will want us to concentrate on thementors' role in identifying external learning resources,legislators will want primarily to see evidence of what ESCactivities cost, while students tend to cite the counselingaspects of this relationship. Multiple frameworks are there-fore a critical part of ESC's research and evaluationactivities.

A third reason for institutional research relates to thesecond: the answers to the questions could have enormousimplications for higher education, Several states, at leasttwo provinces of Canada, and some OECD membercountries have shown interest in nontraditional highereducation ventures. If a large number of ESC students arenotably different from those at "normal" institutions, wecould see many Empire States in the next-few years. if theeducational activities of ESC students and faculty prove ef-fective, we might see their adoption in some form at exist-ing colleges and universities. This will almost certainlyhappen if the cost of these alternative modes of learning isless than traditional modes. An institutional research pro-gram systematically reporting results can facilitate or. if ap-propriate. red-flag such happenings in other places.

Initial Research Strategy at Empire State CollegeThe nontraditional nature of Empire State College

calls for a nontraditional research and evaluation program.Needed is a flexible research strategy which utilizes provenmethodologies of institutional research in combination withappropriate new ideas. Our immediate approach beganwith four major activities that have immediate administra-tive usefulness and that begin to answer to major questionsfacing the institution.

I. Design and Develop Computerized Information Fileson Students, Personnel. Finances, and Learning Re-sources

The geographic dispersion el the College neces-sitates the early placement of visual display computerterminals in the various regior al centers to allowremote input and output. The network will be usefulfor maintaining upto-date records on students and forkeeping the College from becoming buried in paper.The learning resources file will be especially useful forproviding rapid dissemination of knowledge about

65

RESEARCH IMPERATIVE

STUDENTPERSPECTIVE

EXTERNALPERSPECTIVE

(accreditationcommissions,legislators,general public,Bureau ofBudget, highereduccommunity)

66

Table 1

PERSPECTIVES OF EMPIRE STATE COLLEGE

AdvancedStanding

Programof Study

Contract Evaluations

LearningResources

Contracts

Student Mentor

MENTORPERSPECTIVE

INSTITUTIONALPERSPECTIVE

(stated institutional goals &objectives, lop.level administritons goals &objectives,faculty goats &objectives)

available learning resources. For example, the studentinterested in studying the welfare system will be able toreceive instantaneous information about references andlibraries where they are available, internship possi-bilities, lists of knowledgeable people in the com-munity, etc.

2. Winch Case Studies of "Finishers" Persons RecentlyGraduated from the College

This project, which involved intensive in-terviews with ten of the first 30 graduates of ESC andtheir mentors, was recently completed.6 From it, theCollege learned of some areas that need attention and theresearch and evaluation staff learned where to focussome future research efforts.

3. Undertake a Cost/Effectiveness StudyThis three-year project, recently funded, is the

framework into which all other studies will now fall.We currently intend to utilize standardized cognitiveand personality instruments in conjunction with contentanalysis of student learning contract products (e.g.papers), journals and interviews in an unusual rollinglongitudinal design (see Table 2).

The year-round calendar of the College whichallows students to enter throughout the year necessi-tates this design which has several starting points andseveral ending points. Some advantages of a rollinglongitudinal study include: (1) it is more representativeof continuing patterns than a normal longitudinal design

Bradley and Palola

because data is collected over several months at eachadministration; (2) it is more reliable than a normallongitudinal study because tho retestings can serve as acontrol; (3) like any longitudinal study, it allows the re-searcher to follow changes in individuals; (4) the inter-view component increases the likelihood of picking upaffective environmental factors; (5) the design allowsthe researchers to have a large sample without buryingthemselves at any given time; (6) the testretest aspectallows early feedback to the College; (7) the longitudi-nal sample can be used for other studies.

4. Conduct Participant Research at the Regional LearningCenters

The focus of this project was the faculty mentor:What are the mentors' characteristics, activities, prob-lems, anxieties, and ambitions? To find out, we joinedthe regional center staffs for a period of time and es-sentially became mentors. The study also involvedquestionnaires and interviews. An additional majorhope was that through the unusual participantmethodology, the research and evaluation staff woulddevelop a regional center perspective to go along withthe inherent Coordinating Center perspective.

Summary

Institutional research has seldom been an imperativein higher education. However, at Empire State College, adesire for immediate decision-making information, the

Table 2

ROLLING LONGITUDINAL DESIGN

Month

Sample 1 2 3... 18 19 20

A initialTests andInterviews

B

C

InitialTests andInterviews

InitialTests andInterviews

FirstRetests andInterviews

FirstRetests andInterviews

FirstRetests andInterviews

30 31 32

SecondRetests andInterviews

SecondRetests andInterviews

SecondRetests andInterviews

67

RESEARCH IMPERATIVE

amount and breadth of interest expressed in the institution, College. The initial research strategy is to launch studiesand the tremendous potential implications of findings make with immediate utility but which also begin to answer theinstitutional research an integral part of the developing mafor questions facing Empire State College.

Empire Stale College of Slate University of New York. Draft Master Plan 1972, Saratoga Spring, New York, January 1,1973, p. 1.

I Tom Burns and GM, Salker. The Maria-gement of innovation. Tavistock Publications. Ltd., London, 1961.

Paul Lawrence and fay Lorsch, Organization and Environment: Managing Differentiation and integration. Harvard Press,Cambridge, Mass., 1967.

6 Marvin W. Peterson, "Institutional Research and Policy Formulation: A Contingency Model," in Clifford B. Stewart (ed.) institutionalResearch and Institutional Policy Formulation: 11th Annual Forum of the Association of institutional Research. 1971, pp. 27 -32.

Herbert C. Keintan, "Compliance, Identification, and Internalization: Three Processes of Altitude Change," in Proshansky &Seidenberg (eds.) Basic Studies in Social Psychology. Holt. Rinehart, & Winston, New York, 1965, pp. 140-148.

6 Warren G. Bennis, Changing Organizations. McGraw-Hill, Inc., New York, 1966, p. 175.

1 fit B. French and B. Raven, "The Bases of Social Power," in D. Cartwright & A. Zander (eds.) Croup Dynamics: Research and Theory,2nd edition. Row. Peterson & Co., Evanston, Illinois, 1960, pp. 607-623.

Ernest Palola and A. Paul Bradley, Jr., 10130: Ten Case Studies OuI of the First Thirty Graduates, Empire State College, SaratogaSprings, New York, 1973.

68

INSTITUTIONAL RESEARCH AND THE EXTERNAL ENVIRONMENT:AN APPROACH TO CONTEXTUAL ASSESSMENT

At the first general session of the 1972 Annual Forumof the Association for Institutional Research, WarrenBennis took time to reminisce about his experiences asAcademic Vice-President at Buffalo. As hisaddress drew to a close Dr. Bennis directed a few"pointers" toward present and future college and universityleaders. One such pointer he mentioned had to do withknowing one's territory, one's community; that is, knowinghow the institution fits the community, how the institutionis perceived within the community and how the institutionserves and is served by the community in which it islocated. Two days later, on Wednesday, Harold Hodgkin-son admonished us for being reductionistic, that is, for al-lowing our vision to become clouded so that we think interms of components and not wholes. His argument wasvery similar to the one about the forest vs. the trees. Themessage contained in these remarks of the two speakersserves as the introduction for this presentation.

It has always been of extreme importance for collegeand university leaders to have some knowledge of theirinstitution's place in the environment. The words "collegecommunity" more often than not refer to the community ofscholars, students and supportive staff. The College Com-munity is seen as an entity, one with boundaries of varyingdegrees of permeability and fluidity. Today, with publicdisenchantment focused toward higher education, it wouldappear that institutional leaders have a greater than everneed to assess and take cognizance of community percep-tions of and attitudes toward the college or university. Theterm community must be regarded in the broad sense. Forsome institutions, community can mean the world; forothers it can mean a continent or a nation; for others, andthis includes most institutions, community can mean astate, a region, a city or a town. In post-secondary educationwe need to take a broader view than in the past. We need toknow more about how we are perceived and how we areInformally evaluated. One need only briefly scan theliterature of higher education before it is quickly revealedthat colleges and universities have not taken the time northe effort to comprehensively examine their environments.A few institutions have done extensive work along theselines primarily in the areas of institutional financial impacton the region, space planning, program planning,

Paul R. Lyons, Frostburg State College

enrollment forecasting, manpower needs and the like. Onlya handful of colleges and universities have systematicallysought to comprehensively study and examine the institu-tion as it is perceived in the field, in the communities. Youmay be aware of the excellent bibliographies developed bythe Council of Planning Libraries. This organization hascompiled perhaps the best contemporary accounting ofwork completed in the area of general campus communityrelations.

We are wont to make a great deal of assumptionsregarding how the institution is perceived, should be per-ceived or might be perceived. In the absence of any system-atically sought-out data such assumptions are relegated tothe status of good, bad or mediocre guesses. Most of theseassumptions are based upon communication with eitherfriends or enemies of the institution in question. This typeof friendly/hostile informal evaluation can, at times, have avast impact on institutional decision-makers, especiallythose who do not have some form of objective evidence torely upon. Institutional researchers, by design or inclin-ation, have not begun to scratch the surface of the research-able domain of community attitudes, perceptions andknowledge.

The assessment of community attitudes and percep-tions is but one small piece of the ever-changing contextualmosaic of the institutional environment. In a dynamicsociety such as ours, one marked by social pluralism, themosaic as a representation of reality can never becompleted as the perspective is never fixed over time. Theconcep: of contextual assessment must of necessity, be aflexiable, fluid one. This is how contextual assessment isvalued at Frostburg State College, where the Office of Insti-tutional Research has systematically sought to actualize theconcept of contextual assessment, that is, the on-going con-tinual assessment of the institution's environment. Thecontextual or environment assessment approach isessentially a way of identifying the particular elements ofan institution's environment and setting about to assessvarious characteristics of those elements in relation to theinstitution. The elements of the environment for a giveninstitution might include (besides faculty, students, andsupport staff) parents, high school students, the immediatecommunity, the military, the state legislature, the alumni,

69

CONTEXTUAL ASSESSMENT

and so on.At Frostburg State, a period of approximately 18

months saw the identifitation of several environmentalelements along with attempts toward assessment of eachelement. These elements were:

AlumniEntering Freshmen (prior to attending classes)FacultyAdministrative StaffStudents ("natives")Community Residents General PopulationCommunity Residents Positional LeadersTransfer Students

Currently underway is a project which will involve allelements of or a sample of the last six groups representedabove.

The contextual approach to assessment, like anannual giving campaign or an elaborate, multi-facetedmanagement information system takes time to develop,cultivate, test, and evaluate. At Frostburg State we have abeginning, a frame of reference has been established. Dueto staffing and financial constraints the beginning has beenslow and sometimes painful. Thanks to instruments such asthe C.U.E.S., I.F.I., W.I., the task has been made manage-able. Perhaps the most significant virtue of the contextualapproach is that the assessments of various element char-acteristics can be incorporated into the "whole" to give amore comprehensive picture to the viewer. This can beespecially valuable if the basic elements of the environ-ment (students, faculty, staff) are continually beingassessed with regard to program and institutional govern-ance, aspirations, satisfaction, future orientation and thelike. What we are, where are we going can be brought intomuch sharper focus if one can take into account thoseelements of the environment which impact upon the insti-tution. Contextual assessment, then, assists with bothevaluation and planning. This approach is in direct contra-diction of the modus operandi of the reductionist.

The Frostburg State College SurveyAssessment of community knowledge of the college

and assessment of perceptions and attitudes heldconcerning the college were the major areas of inquiry withwhich the survey dealt. It must be pointed out that thesurvey was regarded as a pilot project, one that was to berefined, polished and periodically replicated. Early in theproject the decision was made to include different surveypopulations. First, a random sample of community resi-dents (general population) was to be contacted. Second, anattempt was to be made to contact positional leaders(mayors, councilmen, county commissioners, business andvoluntary association leaders, and the like). As one goal ofthe project was to supply information for decision-making,it was felt that data provided by such an "elite" sample

70

would be at least as valuable as that provided by thegeneral population, In no instance was the sample of"elites" assumed to be representative of a regional or com-munity power structure although it is conceivable thatmembers of this sample may have been members of thecommunity power structure, Due to the fact that this pilotproject started "from scratch," survey instrumentdevelopment was the most costly aspect of the project withregard to time Invested. The survey instrument or interviewguide was a document of 25 items (some being fixed-choiceresponse types, others open-ended) was titled the Survey ofCommunity Attitudes and Perceptions (SCAP). The inputwas provided by the college president and his administra-tive council (academic affairs, student affairs, businessenrols) and the public relations director. The individualsconcerned arrived at a set of questions which they believedaddressed themselves to the major areas of shared andindividual inquiry. The SCAP was revised three times viathe consensual approach before reaching its final form. Inorder to have some basis for comparison among institu-tional and extra-institutional samples, 8 items (questions)taken from the "Meeting Local Needs" scale of the Institu-tional Functioning Inventory (IFI) were included in the 25items, The WI, as developed by the Educational TestingService, Princeton, New fersey, provides a means by whicha college or university can describe itself in terms of anumber of characteristics judged to be of importance inAmerican higher education. Consisting of 132 multiple-choice items, the IFI yields scores on 11 dimensions orscales, each comprised of 12 items. One such scale,"Meeting Local Needs, "is comprised of primarily informa-tional items. Since nearly all faculty and administrativestaff had recently been subjected to the IFI, it was decidedthat the survey populations be subjected to some if not allof the items comprising this scale. Their responses wouldallow for comparison between the survey population and asupposedly highly knowledgeable population (faculty andstaff) with regard to the college. Eight of the si itemscontained sufficient generality to be included in the surveyinstrument (SCAP).

Clear criteria for defining the "community" in whichthe College exists were hard to come by, and any of severalmethods suggested by sociological community study guidesand texts could have been used, but each was arbitrary andeach would have led to different boundaries. Limits werechosen that were no less arbitrary (yet, hopefully, no lessvaluable) than those suggested in the literature: it wasassumed that college/community attitudes and perceptionswould diminish in intensity and accuracy, respectively,with socio-geographic distance from the institution, andchosen as a measure of that distance was the pattern ofresidence of College employees. It was assumed thatemployee commuting distance was a fair measure of theCollege's area of immediate influence. Due to local

geography, radio station signals and local newspaperdelivery have about the same degree of dispersion in termsof area. An inventory of the College Directory was made todetermine the distribution of employees' residences in thearea. Employees' residences were distributed across 15towns and communities. An unstratified random sample ofresidents drawn proportionately (based on employee distri-bution) from these 15 areas was identified (from telephoneand utility company listings). The size of the sample wasbased entirely upon convenience and manageability. Of62,000 inhabitants in the areas selected represented by13,426 households, the total sample population of 234represents approximately 2% of these households. Samplesize had to be kept small since person-to-person interviewswere to be the means by which the data was gathered withthe SCAP instrument. That is, the general population sub-sample (Nrr: 199) was interviewed while the "elites" (N=35,a 70% return)received the instrument through the mail.Trained, although not greatly experienced, interviewersconducted the interviews of the general population over athree-week period. Press releases were issued to legitimizethe survey and interviewers carried identification whichclarified the purpose of the personal contact.

Independent variablesFour independent variables were selected for close

examination. While each variable was not examined acrossall SCAP items, each variable was examined on severalitems. All individuals surveyed were asked a series ofopening general questions which woule. enable theresearcher to "type" the respondent in the different dimen-sions or variables. These variables were:A. Local Cosmopolitan OrientationB. Proximity to Campus (Frostburg residents only

"typed" according to address) Persons in the sub-samplefrom Frostburg (where the College is located) were sub-divided into four categories based upon how close theylived to the campus and to areas of student living orwalking on a regular basis to the campus. The four "resi-dential area zones" were designated as having (1)"immediate proximity to campus," (2) "moderate close-ness to campus," (3) "moderate distance from campus,"and (4) "distant from campus." Some differences wereexpected to emerge among the attitudes and perceptionsof the sub-sa-nples in each residential area.

C. Population (General population) "Elites")D. ,Length of Residence in Area (Long/Short term)Those independent variables that were examined withregard to a given item were examined through the use ofthe Chi Square (Xi) technique.

SCAP Interview Instrument

The content of the SCAP instrument areas of inquirycan be outlined as follows:

Paul R, Lyons

A. General Information (concerning the respondent)These items served to establish the respondent's"typing" for the independent variables.1. Length of Residence in Community2. Local Cosmopolitan Orientation3. Level of Education

B. Information Sources and Opinion1. Sources of Information Concerning the College2. Typology of Information Sources3. Nature of Information4. Perceived Opinions about the College

C. Assessment of Factual Knowledge Regarding theCollege1. Major Differences between the College and Neigh-

boring Community College2. Student Population

a. Size of student bodyb. Permanent residence of students

3. Curricular Offeringsa. Programs offeredb. Future program offerings

4. Faculty and StaffD. Assessment of Attitudes and Perceptions Regar-

ding the College and its Future1. Economic Impact on Frostburg1 Economic Impact on Area3. Recent Growth Rate ("growth" undefined)4. Future Rate of Growth

E. Attendance at College Events1. Attendance Patterns in General2 Attendance Patterns at the College3. Distribution of Attendance across Types of Col-

lege EventsF. Student/Community Relations

1. Frequence of Contact with Students2. Keyword-Descriptions of "Typical" Student3. Major problem areas

G. "Meeting Local Needs" Scale of the InstitutionalFunctioning Inventory

Responses to these eight items were most revealingin that the item content reflected knowledge of Collegeservices, practices, policy and philosophy and compar-isons could be made across samples of the generalpopulation (total sample), "elites," faculty, administra-tive staff, and Frostburg residents. Wide discrepanciesbetween response patterns of community residents andcollege personnel were particularly revealing.

In general, the results indicated that the GeneralPopulation is not aware of 'the nature of servicesthe College is providing or is able to provide. Lackof knowledge characterized the General Populationsub-sample. The Elite sub-sample gave a much higherproportion of "committed" responses (Yes-No) to ques-tions but were highly, almost overly, positive. Facultyand staff, who have more knowledge regarding College

71

CONTEXTUAL ASSESSMENT

functions and services, had similar responsepatterns although staff members were consistentlymore positive with regard to community access toservices and functions.

Summary of Tests of Hypotheses (A-F)Briefly stated, it was a concern of the researchers to

determine which of the independent variables treated inthe foregoing sections had value in terms of identifyingspecific college "publics" each having specific or particularresponse patterns with reference to dependent variables.

Table 1, below, presents the distribution of tests ofstatistical association over independent variablesdichotomized into significant and non-significantoutcomes.

Results indicate that one-sixth (lie) or about 17% ofall tests were significant at the .05 level, Proximity andGeneral Population-Elite variables had a much lowerdemonstration of statistical association with dependentvariables than did Cosmopolitan - Local; Length of Resi-dence variables.

From the point of view of decision-making with an

Table 1

DISTRIBUTION OF OUTCOMETESTS OF SIGNIFICANCE

IndependentVariables

OutcomesNon-Significant Significant

N I % of Tests I N I % of Tests

Cosmopolitan-local 7Proximity 7

Gen. Pop.Elite 13Length of Residence 3

701009360

3

1

2

30

740

Total Tests 30

% of all Tests 83%

617%

eye to the future and to various college publics, thisfunding presents a problem: it is far easier to identifypersons according to Proximity and/or inclusion in theGeneral Population or Elite (positional leaders) samples,but these groupings, unfortunately, do not demonstratestatistical linkages with independent variables. On theother hand, Cosmopolitan-Local orientation and Length ofResidence variables represented by population elementsare much more difficult to identify yet a higher degree ofstatistical association is eveldent with regard to particulardependent variables.

Because the results in terms of total number of testsof significance are somewhat dependent upon the parti-cular matching of independent-dependent variables fortesting, it seems reasonable to conclude that no realpatterns of association are discernible from the presentdata. Subsequent follow-up studies will, hopefully, develop

72

more sophisticated means of identifying specific "publics."

Survey Results and Their ApplicationWhile reporting specific findings is perhaps

inappropriate for this presentation, some indication of howthe survey results were applied at the College may bemeaningful for researchers. What follows represents someof the effects the survey has had on the "Collegecommunity."1. Enhanced Awareness and Knowledge of Top-Level

Decision-MakersAlthough the instrument was not sophisticated, a

great deal of information was gathered which, hereto-fore, was unavailable. Some commonly held stereo-types regarding the community were shattered and theavailable information made it possible for one tovisualize how the "typical" community resident views

4

the College. In addition, information was availableindicating what some of the future directions of theCollege might be

2. Public Relations/Information Services ReEvaluationSince the results of the survey were made avail-

able in May of 1972, the Public Relations staff at theCollege has initiated a series of activities to moreadequately project what the College is and is not and,more importantly, the approach taken in informing thepublic has been changed dramatically.

3. Survey Report as "Required Reading" in Some Depart-ments

Some administrative and academic unit personnelwere asked to closely examine the content of thesurvey analysis with the expectation that said contentwould be of value in evaluation and planning types ofactivities.

4. Long-Range PlanningAs the College will continue to assess many

environmental elements, the community survey (or asimilar type of assessment) will remain as one of themajor or basic elements for gathering information,Survey results have been useful in the long-range plan-ning effort of the public relations I informationservices staff, the results are being scrutinized by anad hoc committee which is setting goats (long-range)for the Co 'ege's graduate program, and they haveserved as reference material for the College's academicplanning group (undergraduate program).

Paul R. Lyons

A study currently underway at the College ismaking use of the Institutional Goals Inventory (I.G,L)developed by the Educational Testing Service,Different college constituencies (and external groups)are queried by the 1.0.1. instrument to seek to establishthe importance of both process and output goals. Acommunity sample has been selected for inclusion inthis study with the anticipation that not only willcommunity residents' responses be valuable for thecurrent purposes of the investigation but that theirresponses may serve to test the reliability of theprevious community survey.

6. Follow-upThere were enough "surprises" contained in the

results of the survey to warrant further and differenttypes of inquiry. In addition, these same results havehelped to create a more open, inquisitive attitude onthe part of College officials with respect to what is "outthere."In brief summary it can be said that the O.I.R. at

Frostburg has sought to "know the territory" as WarrenBennis had suggested. It is believed that any college oruniversity that makes an attempt to develop informationabout their community or sphere of influence will be light-years ahead of those who are content to play guessingga e'11, s and who are confident that they "know" whatpeo le,think. Without systematic assessment and inquiry itis easy to "feel" but not to know.

73

PLANNING MODELS IN HIGHER EDUCATION:A COMPARISON OF CAMPUS AND RRPM1

Over the past decade, higher education has experi-enced a painful transition, from a growth period in whichpost-secondary education enrollments more than doubledin size, a wave of traumatic student unrest and rebellion,and, now, a phase of consolidation and retrenchment aperiod of sober re-evaluation of the purposes and contentof higher education.

The difficulties of the past decade have given rise toa massive urge to apply the contemporary technologies ofmanagement science, operations research, modeling, andsystematic decision theory to the planning and manage-ment processes of colleges and universities. The attempt toapply rational economic models to the complex politicalsystems of governmental and social institutions has fallenfar short of demonstrable success in relation to our hopesand expectations. Nevertheless. the search persists for moreeffective ways of utilizing the resources of higher educa-tion.

Large-scale computer models are one manifestationof the "managerial revolution" in higher education. In a1970 survey by Weathersby and Weinstein, 31 highereducation models were identified and compared.3 Amongthese, eight were comprehensive institutional models butonly two were then operational CAMPUS and the Uni-versity of California Cost Simulation Model (CSM).4 Sincethen several other model systems have become operational.These are: RRPM, CAMPUS/HEALTH (a special version ofCAMPUS for medical schools). and the TULANE model, Aset of related modeling systems are CAP:SCISEARCH andHELP/PLANTRAN. The former is designed primarily forsmall liberal arts colleges, while the latter is mainly abudget simulator. There are a number of models being de-veloped for a particular institution, such as RCN andFACSIM, specialized models for the Air Force Academy.

However, for the many institutions of highereducation that cannot develop models of their own and areseeking a generalized and comprehensive model, there aretwo that have attracted greatest attention: RRPM, the mostaccessible and commonly used and CAMPUS, the most de-tailed and comprehensive. These two models are thesubject of this paper.

After a brief historical review of the models, discus.sion will cover the basic structure of the models; common

74

KM. Hussain, New Mexico State Universityand Thomas R. Mason, University of Colorado

elements and differences between them; their implementa-tion and use; and finally, some comments on some de-sirable features for their further development,

HISTORICAL DEVELOPMENTCAMPUS is an acronym for "Comprehensive Ana ly-

lice! Methods of Planning in University Sustems." It has itsorigin in the work on simulation in higher education doneby Judy and Levine.5 They developed and now marketCAMPUS through the firm, the Systems Research Group(SRC) based in Toronto. Judy and Levine developed CAM-PUS V6 under a grant of $750,000 from the Ford Founda-tion, which was placed in the public domain in 1970.

CAMPUS V was hardly used. It was inadequatelydocumented and very costly both development andoperational costs. The development costs were highbecause of the large mass of data required by the model,including resource data on each activity a set or subsetof a course that requires a unique set of types of resources.These were used for computations done mostly L, one inte-grated main program that led to high operating costs, es-pecially when answering "What if" questions in ti.e simul-ation mode. Also, most of the data had to be kept in resi-dent core requiring a very large computer. For the Uni-versity of Illinois, this was estimated at 4 million bytes,recently reduced to 300,000 bytes as a result of much repro-gramming. Thus CAMPUS V was beyond the reach ofalmost all institutions except the large and the daring.These included the State University of New York at StonyBrook and the University of Illinois. The State of Minnesotaalso implemented CAMPUS V on a pilot basis (in oneschool of a univeristy, one state college and one juniorcollege).' Despite its limited use, CAMPUS V did performan important service to higher education. It demonstratedthe feasibility of a comprehensive cost simulation modelthat could improve decision-making in planning andbudgeting. What was needed, however, was a model thatmade more modest demands on data equipment, andanalytical effort so that it could be within the reach of mostinstitutions of higher education,

To achieve such an objective, the U.S. Office ofEducation funded a proposal for model development by theNational Center of Higher Education Management Systems

(NCHEMS) at the Western Interstate Commission forHigher Education, This product is known as RRPM-1 Re-source Requirements Prediction Model.

RRPM 1.2 was the first operational version. It was amodification of the California CSM0 made by the staff ofNCHEMS and a national task force. II was implemented byeight pilot institutions selected to represent the differenttypes and sizes of institutions in the country. As a result ofthe pilot testing,fulther modifications were made andRRPM 1.3 was released in mid-19710.

Since RRPM 1.3 was a generalized model, otherspecialized versions were planned. Versions 410 and 5 wereto be specialized for the community colleges and the statecolleges respectively. But instead a sixth version, RRPM 1.6was released. This invoked reprogramming and rearrang-ing of input data by discipline rather than function. Thisdecreased core requirements (along with the fact that it

stores only non-zero data) in spite of the relaxation of con-straints on the dimensions of student programs and dis-ciplines. This made its use independent of the type of insti-tution using it. It is also conceptually simpler than RRPM1.3 because it has no space-management module and hasfewer relationships for support costs. II was implementedin 197211 and has now been released to the public. Bothversions 1.3 and 1.6 are now operational, but only 1.3 isbeing maintained by NCHEMS.

Meanwhile CAMPUS underwent considerablechanges. It was completely reprogrammed as CAMPUS VIand was made modular. This greatly reduced its opera-tional costs and core requirements. But data are still re-quired at the activity level. However, the input formatswere changed and documentation was greatly improvedmaking data preparation much easier. CAMPUS VI was re-programmed as CAMPUS/COLORADO and 12 CAMPUSVII, making them more modular, more flexible in their di-mensions, with additions in the costing routines, and abetter handling of the research :ector.

In addition. CAMPUS VII was developed.ls Thisversion does not require data at the activity level and hencehas further reduced core requirements and operationalcosts. It is designed for institutions requiring data only atthe aggrevated level of department or above.

In summary, there are two operational versions ofeach family of models worth examining: Versions 1.3 and1.6 for RRPM; and Versions VII and VIII for CAMPUS. ButRRPM 1.3 and CAMPUS VIII are conceptually the morecomprehensive of their family of models and as such willbe the basis of most of the discussion that will follow. Dif-ferences in the other operational models. when significant,will be so identified.

Basic LogicThe basic logic common to both RRPM 1.3 and

CAMPUS Vill is shown in Figure LH There are many

Hussain and Mason

minor differences that do not appear in Figure 1 and havebeen deleted in order to keep the figure simple. Anexample is the computation of instructional load, In RRPM,this is calculated in credit hours and is then converted tocontact hours by a ratio of contact hours to credit hours.This conversion is not required in CAMPUS which has allits activity loadings expressed in contact hours. it is alsounnecessary In RRPM if the student loading is done in con-tact hours to start with. There are many other such minordifferences that will not be discussed. There are, however,two main differences in the logic of RRPM and CAMPUS.They concern the level of detail in instructional loading aswell as in planning factors. These will be discussed in turn.

The instructional loading in RRPM is done throughan induced course load matrix (ICLM). In this matrix, eachcolumn represents the credit hour load induced by astudent major (at different levels of academic achievement)on different levels of courses that are offered by differentdisciplines. In CAMPUS, the load induced is not in termsof aggregated levels of courses (as in RRPM), but in termsof specific courses or activities. Examples of activitieswould be a course with two weekly contact hours of lectureand one weekly contact hour of lab (or recitation). Sincethey require different types of resources (personnel andspace), the lecture and lab are separate activities. The de-tailed level of activity does generate a variety of reports thatcan be very valuable, especially in costing which will bediscussed later. it also enables planning at the mostelemental academic organizational level and involves allorganizational levels in the planning process. However,there is a price that must be paid: the massive detailed datainput required at the activity level.

In the case of the University of Colorado, there areover 2,000 activities and for each activity up to 16 data ele-ments on resource loading have to be specified. These datamust be collected (typically on forms) converted tomachine-readable form, stored, processed, and maintained.The maintenance cost (especially on the mix of activitiesrequired for each student program) could be high for in-stitutions where student preferences change or wherecourse requirements change significantly. It is difficult topredict in cases of new student programs and degrees. It iseven difficult to specify the activity mix in order to main-tain status quo. This is due to the fact that the student loadand activity mix varies not only between semesters but alsoamong types of programs such as daytime and evening pro-grams. The mix can be unstable even at the aggregatedlevel of the ICLM, as was experienced by the pilot institu-tions of RRPM 1.315 and other ICLM studies.16 These studiesshow that the greater the disaggregation of the ICLM, themore the instability. This instability will increase asstudents demand and get more electives and unstructureddegree requitements. This will increase the problems ofpredicting new course mixes and the redistribution of old

75

CAMPUS AND RRPM

Table 1

SOME PLANNING FACTORS IN RRPM AND CAMPUS

INSTRUCTIONAL

Student programs

Faculty

SPACE

76

CAMPUS VIII RRPM 1.3

by detailed courseactivity mix

resource loadingof each activity

spacepersonnel typetime of offering

section sizeaveragemaximumminimum

maximum number ofsections

substitution policycontract lengthturnover rates andhiring policysabbatical policyweekly availabilitypromotion policyaverage salariesby rankrank distributionacademic levelworkload weightsadministrative load

substitutionpoliciesavailabilityutilizationtypesizeconstructioncoefficients

by mix of credit hoursin department/discipline

loading of groups ofcourses at differentlevels and fields

spaceequipment

section sizeaverage

average salaries byrankrank distributionacademic level

(not in RRPM 1.6)

availability

typesizeconstructioncoefficients

ones that are dropped.The detailed level of data required by CAMPUS

occurs throughout the model and is reflected in theplanning variables. As an example, consider the determina-tion of the number of sections required. In both the RRPMand CAMPUS, the number of sections is determined bydividing the student load by the average section size andusing some rule for accounting for the leftovers. InCAMPUS, however, the solution is subject to many con-straints such as maximum class size, minimum class size,and maximum number of sections. This adds to the controland flexibility that the user has, but it requires that allthese constraints be specified as planning variables (one setfor each activity).

For some institutions this choice of planning vari-ables is often unnecessary, and this has been recognized inRRPM 1.6 where the faculty FTE can be calculated by anoption that uses a weekly credit-hour (or contact-hour) toadby level of course, thereby eliminating the planning factorsof average section size, credit to contact-hour ratio, distribu-tion of contact hours, distribution of faculty rank, and aver-age faculty work load by rank.

There are other planning factor differences betweenRRPM and CAMPUS. These are listed in Table 1. One setfound in CAMPUS alone enables the "flowing" of facultybetween time periods. using rates of turn-over, sabbatical,and promotion policies, contract lengths, and availabilityperiods. CAMPUS maintains a faculty inventory for eachtime period. It also allows for substitutions among rankswithin a cost center but not any substitution among disci-pline specialities within ranks.

There are three parts of the CAMPUS VIII modelother than those shown in Figure 1. One part is a StudentFlow module that goes in front; one part calculates non-teaching-salary costs; and a third part is a costing module.Each will now be discussed.

The Student Flow ModuleThis module determines the student enrollment in

each student program at each level of student academicachievement (freshman, sophomore, etc.). It is part of theCAMPUS VIII package and determines the flow of studentsthrough the system by using pass-fail rates at each level,repeat rates at the same level, dropout rates at all levels,and transfer rates between programs. This is conceptuallysimilar to the Student Flow Model developed by NCHEMSthat is designed to interface with RRPM." Following theNCHEMS tradition this model was developed by its staffsupported by a national task force, tested at selected pilotinstitutions and implemented successfully."

The NCHEMS student flow model and the one inCAMPUS VIII have much in common: both can be by-passed if desired; both use data on freshmen enrollmentand transfers as exogenous variables; and finally, both haveproblems and issues raised by using the transition matrix.Some of these issues include: the definition of points and

Hussain and Mason

student states most suitable for the transition matrix; thecalculation, aggregation and stability of transitionalprobabilities; and the validity of the Markovian assumptionfor student transitions. These issues are part of the ongoingresearch and development work being done at NCHEMS.

Non-Salary CostsNon-salary costs are calculated in both RRPM 1.3

and CAMPUS VIII at the cost center level. This requires theestimation of both the relationship and the cost coefficientsat the cost center level. This is no trivial task. In the Univer-sity of Colorado, implementation of CAMPUS requiresstating 2,800 equations and estimating cost coefficients.Over 43 variables are used in these relationships, most ofthem being endogenously determined. (CAMPUS VIIIallows up to 130 such variables, and 13,000 equations.)

CostingBoth CAMPUS and RRPM calculate costs for aca-

demic and support program and subprogram levels of theNCHEMS Program Classification Structure. In addition, inCAMPUS VIII, the costs are aggregated by budget functionand object category. This facilitates preparing annual line-item budgets for financial control needed in addition toprogram budgets for analysis and decision making.

Both CAMPUS and RRPM calculate unit costs forstudent programs by contact hour, credit hour and FTE fordifferent levels of aggregation. In addition, CAMPUS calcu-lates the direct cost for each activity. This is aggregated foreach activity in the activity mix of each student programand gives annual student program costs. In RRPM, thestudent program cost is determined as the inner-product ofaverage cost per student credit hour by discipline trans-posed back through the ICLM to student program. Theaverage cost figure, however, lay result in student pro-grams using less than average cost courses in the disciplinebeing overpriced and programs using higher than averagecosts being underpriced. This possibility does not occur inCAMPUS because of its detailed activity level costing data.

Indirect costs are allocated to primary programs (In-struction, Research and Public Service) in both CAMPUSand RRPM 1.3 but nut in RRPM 1.6 because this is to bedone in the Cost Finding Principles Program of NCHEMS.9II has software that will allocate support costs to primaryprograms and can be used as a costing module inde-pendently or in conjunction with RRPM.20 The project soft-ware could also be used in a simulation mode to experi-ment with parameters of allocations. Once the parametersare selected they can then be used for allocation in RRPM1.3 or 1.6. In CAMPUS, there are plans for options as tosome allocation rules: by a specified percentage; in propor-tion In the direct cost of the receiving categories: or 8con.hi notion of those two rules.

The Cost-Finding Principles project is Oro expectedto suggest procedures for cost exchange amonti institutions,another of NCHEMS projects."

77

CAMPUS AND RRPM

Cost allocation raises numerous problems includingone of allocating faculty effort between instruction, ad-ministration, research, and public service. Should thisallocation be done by assignment or by actual effort distri-bution? if the latter, how is faculty effort to be measured?This problem has been a concern of institutional research-ers for over a decade and has not resulted in much agree-ment. For example, in measuring faculty effort, 24 studiesmeasured hours spent weekly, while 16 used percentagedistribution of time, and four studies used both. 22 Thisproblem and related ones, ax the subject of yet another

78

NCHEMS project: The Faculty Activity Analysis.23

Differences between RRPM and CAMPUSMany differences between RRPM and CAMPUS

have been identified above. Other differences are in the di-mensions of the model. These are shown in Table 2. Somedifferences of a technical nature are listed in Table 3. Othermain differences concern the revenue module, capitalbudget, output reports, and implementation considerations.These are discussed below:

Table 2

A COMPARISON OF DIMENSIONS*

CAMPUS RRPM

VII VIII 1.3 1.6

Student Programs 20 350 90 200

Academic Disciplinesor Departments (Teaching

30 100 90 200

Cost Centers)

Nonacademic Departments 10 25(Nonteaching Cost Centers)

Activities 0 4000 0 0

Course Levels 1 (implicitin activity

4 7

specifica-tion)

Instruction Type 3 9 4 S

Ftudent Levels 4 8 7 7

Faculty Ranks 5 10 S 6

Nonacademic Ranksor Classifications

7 150 4 4

Space Type and Size Rangesacademic 8 125 2 0nonacademic 10 110 4 0

NonpersonelResource Types 7 120 3 7

'Source of data:A.P. Van Wijk and R.S. Russell, "State of the Art in Educational Cost Modelling." Paper presented atthe ORSA, TiMS, and AllE Joint Meeting in Atlantic City, November 1972, Toronto, Canada: SRG.,1972, p. 35; K.M. Hussain, op. cit. p. 18, and D.G. Clark, el al., introduction to the ResourceRequirements PredirOon Model 1.6, Boulder, Colorado, NCHEMS at WICK 1972, p.6.

Revenue ModuleRRPM has no revenue module, while both CAMPUS

VII and CAMPUS VIII do. In CAMPUS, revenue fromstudents is estimated as a function of prejected enrollmentand tuition rates. Revenue from public funding agenciesis calculated by formula which in many cases must be re-stated and reprogrammed in order to meet local needs. Therevenue components can be projected from year to yeareither by an absolute value or by a given percentagechange. The model does not include important componentssuch as financial aid and portfolio management." What itdoes include is grants. gifts. endowments. and special reve-nues which are treated exogenously instead of makingthem a function of endogenous variables such as studentenrollment number, type of student programs, etc.

Capital BudgetBoth RRPM 1.3 and CAMPUS VIII calculate the in-

cremental cost of capital expenditures resulting fromprojected increase in space requirements. To calculate this,however, CAMPUS VIII has a greater facilities-planningcapability. It "shuffles" rooms around according to givenspace substitution and utilization policies: calculates netshortages and surpluses of space by type: calculates spaceutilization and maintains inventories of rooms by size andtypes.

CAMPUS VII calculates the square feet of space andnumber of stations required for its eighteen space types.RRPM 1.6 has no space management or capital budgetingcapability.

Output ReportsAll operational versions have sets of output that vary

in number and content, but CAMPUS VIII has by far themost detailed and comprehensive set. Its input data, beingcollected at the lowest activity and organizational level,enables it to aggregate its output at all organizational levels.The output is particularly good for space management andon administrative indices on loading, costing, andutilization.

RRPM 1.3 has a unique report. It identifies on onepage the results of len sets of changes made in the "whatif experimentation mode." The display of results faci-litates an analysis of incremental changes and sensitivityanalysis. For such experimentation the user may makeblanket changes by percentage or an absolute value in addi-tion to replacing one or more values. CAMPUS and RRPM1.6 allow only a percentage blanket change and only for aselect set of variables.

RRPM 1.3 also has a TRACER - TRAINER routinethat "traces" all the intermediate output for any one se-lected discipline." This is useful not only in training a useron how the model handles his data but also in debuggingand in validating the model.

An institutional implementation of RRPM 1.3 has the

Hussain and Mason

TRACER on a terminal in the programmed instructionalmode along with routines to help the uninitiated user.CAMPUS also has a CAI package. It has an interactiveprompter which is especially helpful to the user.

One final set of comments on comparing differences:one must recognize that RRPM has always been con-sidered by its designers as one component of an ongoingprocess of developing an inter-related system for planningand management in higher education. The NCHEMSStudent Flow Model, Cost-Finding Principles, OutputEvaluation, Facilities Analysis and other projects inter-relate with RRPM as part of the larger planning andmanagement system.

Perhaps the most important NCHEMS project is thaton outputs of higher education. This project is concernedwith identifying the output" and developing instrumentsfor measuring them. Once this is done, then it will bepoisible to conduct cost-benefit analysis which is so im-portant to evaluating performance. This capability will alsoeliminate the current danger and concern that unit costingwill be misused by funding agencies for allocating re-sources among institutions. There is the danger, as CaryAndrew points out, that "without definitions and measuresof outputs, the cost simulators are likely to lead adminis-trators to take unwarranted actions . . . Cost simulatorswill most likely get blamed for some of the intermediatebad decisions and not get credit for speeding up on ourway to an improved management process by forcing us to'get off our hands and counter the threat.' "26

Cost of ImplementationIn evaluating models, one should look at their cost-

effectiveness ratios. The effectiveness of a planning model,however, cannot all be quantified. Its implicit value could,however, be compared to costs, and a judgement could thenbe made as to whether or not the model is worth the cost.

In calculating the costs, one must differentiate be-tween development and operating costs. Development costsinclude the cost of software discussed in Table 3. Other de-velopment costs include costs of changing the model tomeet institutional needs, validating the changes, datageneration, and training. Typically. the largest componentis that of data preparation. This, in turn, depends largelyupon the type and quantity of data required. This could becompared for actual implementations of the models, butsuch comparisons invite suspicion since institutions to becompared are often different in structure and complexityand, have different data bases. To overcome this problem,one could compare the data generation problem for differ-ent types of institutions for each of the models. This isdone for four types of institutions whose institutionalcharacteristics are compared in Table 4. Their data genera-tion problems are compared in Table 5.

Data generation is important not only in estimatingdevelopment costs, but also in estimating operational costs.

CAMPUS AND RRPM

80

Table 3A COMPARISON OF SOME TECHNICAL DATA

CAMPUS RRPM

VII VIII 1.3 1.6

t.

Program Language Used FORTRAN FORTRAN FORTRAN COBOLIV IV & (ANSI)

COBOL

Equipment Used Most Com- IBM IBM IBMputers CDC CDC CDC

upward ofan IBM

UNIVAC UNIVACBORROUGHS

1130

Minimum Core 16 256 128 50 for DOSRequirements(thousand bytes)

65 for OS

Cost of Softwarepurchase $12,500 $25,000 $50.00 $50.00lease $ 3,000 S 6,000

+ 350/mfor 36m

+700/mfor 36m

Consulting services for NCHEMS provides limitedoverall project manage- training at nominal cost.meet, training of senior Var es Other help in implementationstaff and adaption of $5,000 to $50,000 is a function of the supplyplanning manuals. and demand of its staff.

Source of data:SRG: "A State of the Art review of Computer - Based Educational Planning Systems," DRAFTUnpublished, Aug., 1972, pp. 48, 50, 51, and A.P. Van Wijk and R.S. Russell. op. cit., pp. 32.35 onCAMPUS; K.M. Hussain, op. cit. on RRPM 1.3, and Clark, et al, op. cit., p. 31, RRPM 1.6

Hussain and Mason

Table 4

CHARACTERISTICS OF INSTITUTIONS COMPARED (FOR 1971-72)

v u t - am .tg n vers ty tateCollege

Commun tyCollegeMain Campus --.illranch Campus

Student FTE 18,632 2,283 5,284 2,400(Average for Year) (Semester) (Semester) (Quarter) (Quarter)

Highest Degree Offered Ph.D. Masters Masters Associate

Student Programs 275 26 65 58

Cost Centers - Academic 53 12 46 11

Cost Centers - Non-academic 47 9 20 21

Academic Activities 2,200 724 1,200 244

Instruction Types 9 6 4 2

Student Levels 8 6 6 1

Course Levels 5 5 5 1

Faculty Rank 9 5 8 1

Non-Adacemic Rank 38 15 33 5

Space Types 70 29 15 15

Other Resource Categories 19 10 10 3

Other Resources Subcategories 56 31 4 7

81

CAMPUS AND RRPM

Table 5

COMPARISON OF COSTRELATED ELEMENTS

Multi-campus University Stale

College

Community

Col egeMain Campus Branch Campus

DMA Elementsin ICLM (in 000's)

CAMPUS RRPM CAMPUS RRPM CAMPUS RRPM CAMPUS RRPM

VII VIII 1.3 1.6 VII VIII 1.3 1.6 VII 1 VIII 1.3 1.6 VII VIII 1.3 1.6

18 1815 18 18 .3 19

4

.3 3 78 3 3 .6 14 .6 .6

Regression Equationsfor no/1,401nssalary costs

2800 300 610 2400 600

Development Costsin 000's of S

12 80 15

4

9 4 20

.25 5 10 26.4 32 7 4 20.7 24 6

Run Cost in S"for one completeset of reports

3 SOO 3 2 200 2 2

-.-

300 3 2 200 2

Maintenance 'Cost in Analystfit

1.0 3.0 1.5 1.0 0.25 0.5

4

0.75 0.5 0.75 1.0 0.5 0.75 0.5 0.5 0.25

'these costs du rut ins Jude the tug eel a dmini.frilor. noilinn on for model, odof softhate.sonsultinx or us ethead Sume eel ends be sunk t ost.

"these togs need nut be uutol,put togs. epet lairs ti the model hrtiome.peel of the 'normal" operations of the instiluhun

Such costs increase greatly when data elements are updatedannually. In a cost study done for CSM (the conceptualbasis for RRPM), Hopkins found that annual updating ofthe data base more than doubles the annual operating costsof maintaining the model.29 Maintenance of the data base isnecessary in order to reflect the changing values of many ofthe parameters in the model. In a study of some depart-ments in Berkeley, Breneman found that the faculty re-quirement coefficients vary as much as 200% from one yearto the next. x,

Changes to parameters also result during experi-mentation when making simulation runs. This increasesthe operational cost (the cost of each simulation run isshown in Table 5), which is a necessary cost if one wantsto investigate the consequences of possible changes.

Other components of operational costs result fromthe need for continuing analysis of output, and the need fortraining the user. These components are very importantaspects of implementation and arc not sufficiently recog-nized. The estimates for this effort is shown in Table 5.11will vary with institutions and is a function of theirplanning experience; the support that they can get fromother departments such as the Computer Center; thenumber or nature of modifications to the model initialed bythe user and the Computing Center; the extent of the use ofthe model: and finally the thoroughness with which thetask is performed.

82

In Table 5, the figures for CAMPUS VIII are basedon emperical data. The figures on RRPM were calculatedfrom the manual on that model.3t The figures on the re-maining models were estimated by the writers and theircolleagues who are knowledgeable about the institutionand very knowledgeable about implementation of themodels concerned. The figures were then checked againstpublished data on implementation. Unfortunately there arenot much data on the implementation of RRPM 1.6 andCAMPUS VII since these arc relatively new models andhave few implementations. The number of implementationof these and other models are shown in Table 6.

Summary and ConclusionsIn comparing the class of operational models of

RRPM and CAMPUS, one can identify the models thatoperate at the discipline or department level as beingRRPM 1.3, RRPM 1.6 and CAMPUS VII. Among these,RRPM 1.6 is conceptually the simplest and least compre-hensive. It is also the cheapest in both development andoperational costs. RRPM 1.3 is slightly less comprehensivethan CAMPUS VII but is cheaper to develop largely be-cause of its negligible costs of software. 11 does, however,require more computer core memory than does CAMPUSVII or RRPM 1.6

CAMPUS VIII is the most comprehensive of all themodels examined. It is also the most detailed both in the

Hussain and Mason

Table 6

NUMBER OF IMPLEMENTATION*

CAMPUS VII 3

CAMPUS V, VI £ VII 36

RRPM 1.3 70RRPM 1.6

CEM 110

Source of Data: for CAMPUS, the data came from A.P. Van Wilk and R.S. Russell Resk, p. 34, and for RRPM and CEM the data came from NCHEMS at WICH E.

CEM is a training version of R4PM that after being used for training is also being usedfor planning in the operational mode. It is conceptually similar to RRPM 1.3, hassmaller dimensions than either 1.3 or 1.6, and has been implemented only on IBMequipment.

"These figures are fir programs distributed, not necessarily all implemented.NCH EMS does not implement or control the use of its software and hence has no wayof knowing exactly how many of its programs have actually been implemented.

input required and the output produced. It is, therefore.more suitable for decision making at the detailed and de-partmental level (for budget or curriculum planning), butthe price of such capability is larger core requirements andhigher costs of both development and operations. It alsohas the longest lapse time for implementation (between 9-12 months as compared to 2-8 months for the others).

Both CAMPUS and RRPM in all their operationalversions have some common characteristics: they are costand resource models, not cost-benefit models; they aresimulation models, not optimizing models;31 they are sub-system models, not total system models; they have mostlylinear equations for calculating their non salary costs anduse marginal costs. thereby ignoring economies of scale anddiscontinuities; both are basically student-driven; neithermodel predicts the number of new entrants to the insti-tution nor do they relate it to manpower requirements; andfinally, both are deterministic models (except for the prob-ability matrix used in the student flow module).

From the viewpoint of helping the user implementand use the model, neither family of models provides helpin formulating the support (non-salary) cost equations norin calculating the cost coefficients. Also, no help isprovided in studying and improving tLe stability of para.

meters. especially the ICLM. Some work has been donewith CAMPUS and RRPM 1.3 in using terminals but notenough work done on the economics and feasibility ofusing the model to respond to "What if" questions in theon-line-real-time mode. Also. no help is provided to theuser in searching through the very large set of permutationsof possible alternative strategies. Search routines for identi-fying "promising" and near optimal strategies will greatlyhelp the user. Routines that will "bound" the feasibleregion will help also. The current output will help the userif it were packaged with graphics that show "trends" and"gradients" rather than masses of numbers. Reports shouldbe designed that also help in management by exception byidentifying infomation and variations that exceed allowablelevels. Finally, neither model enables the user to calculatetrade-offs directly. For example. if one wishes to find thetrade-offs between average section size and faculty loadwhich keep the cost constant, one has to guess at pairs ofvalues, calculate the costs, and then plot an iso-cost curve.This can be both costly in computer time and slow in re-sponse time.

Some of the limitations discussed above will undoubtedly be relaxed in future versions of CAMPUS andRRPM (and related models). The RRPM development is

83

CAMPUS AND RRPM

related to projects on Identification and Measurement ofOutcome of Higher Education; Faculty Activity Analysis,Program Classification Structure, Data Element Dictionary,Cost Finding Princfple-s, and Information Exchange. Forimplementing these models. NCHEMS is federally fundedand can call upon the advice and help of personnel fromany of its participating institutions. All these resources willbe required to implement the many ambitious projects thatNCHEMS has planned. In doing so. it is hoped that it willnot "drive out" organizations in the private sector thatdevelop models such as CAMPUS, HELP/PLANTRAN. andSEARCH. These models provide an important choice tomodel users; new users, as well as some that have imple-mented RRPM and want more detailed planning informa-tion, To the latter group, CAMPUS VIll is a logical choiceprovided they can afford the extra cost. As long asCAMPUS continues to remain compatible with basicWICHE products on definitions, classifications, and mea-sures, it will offer an important upward compatible choice

for users that have had their planning model experiencewith RRPM.

Institutions for which neither RRPM nor CAMPUS isadequate such as institutions with heavy noninstruc-tional emphases may still find that CAMPUS or RRPMprovide a basis for building an institutionally uniquemode1,33 These models also provide an important trainingground for use of planning models. The benefit of theselarge-scale models for structuring and integrating the insti-tution's basic information system will be discussed inMadelyn Alexander's paper being given at this Forum onimplementation of CAMPUS.34

Finally, the implementation of these deterministiccost simulation models may be expected to lay the founda-tion for development of the next generation of planning-budgeting models. Hopefully, this next generation ofmodels will provide heuristic and optimizing capabilitiesand will incorporate benefit values as well as costs.

Both authors are involved in implementing CAMPUSICOLORADO (a version of CAMPUS VIII) on campuses in Colorado. Also bothauthors were on the Task Force that designed RRPM 1.3. One author was responsible for one pilot implementation of RRPM 1.2.

t See F,1. Lyden and E.G. Miller, Planning, Programming, Budgeting, Chicago: Markham Publishing Co.. 1968; and J. Farmer, WhyPlanning, Programming, Budget Systems far Higher Education?, Boulder. Colorado: WICHE. 1970.

3 G. Weathersby and M. Weinstein, "A Structural Comparison of Analytical Models for University Planning." Paper. P-12, Berkeley,Calif.; Office of Analytical Studies, University of California, Augusl. 1970. For a similar study. see J.A. Casico. "Planning Techniques forUniversity Management, A report for the American Council on Education." Spring. 1970. In ii Casco present; 30 models of the 40 reviewed.See also C. Lovell. Student Flaw Models: A Review and Conceptualimlion, Boulder, Colorado: NCHEMS at WICHE, 1971.

4 References for the source do.mments on this model and others mentioned on this page can be found along with annotatedbibliographies in K.M. Hussain. A Ruourco Requirements Prediction Model: Guide for the Project Manager, Boulder, Colorado: NCHEMS atWICHE. 1971, pp. 94.108.

s R.W. Judy an.] J.B. Levine. A Tool for Educational Administrators, Toronto: University of Toronto Press, 1965.

* For a discussion of the development of CAMPUS 1.1V. see RM. Judy, "Systems Analysis for Efficient Resource Allocation in HigherEducation" in J. Minter and B. Lawrence (eds.) Management Information Syslems: Their Development and Use in Higher Education,Boulder. Colorado: WICHE. 1969. pp. 41-58.

Project PRIME, under the direction of Gary M. Andrew. produced a set of 16 reports by different authors, Minnesota Higher EducationCoordinating Commission. 1971.

B C. Weathersby, "Development and Application of a Cost Simulation Model," unpublished monograph, June 15. 1967. For a critique ofCSM see D. Hopkins. University Cast Simulation Models: A Critique, Berkeley. University of California. 1969. For a response to the critique.see Gary M. Andrew "Further Thoughts on the Use of Large-Scale Simulation Models for University Planning," Unpublished paper, Boulder,Colorado: March. 1971.

See K.M. Hussain and 1. Martin. A Resource Requirements Production Model (RRPM Report on the Pilot Studies, Boulder.Colorado: NCHEMS at WICHE. 1971.

101(11PM 1.4 was implemented at the New Mexico Juntas College according to the specifications made by the Ad Hoc Commitlee on thedesign of RRPM 1.4,

11 R. Huff el al., Implementation of NCHEMS Planning Models and Management Tools at California State University, Fullerton,Boulder. Colorado: NCHEMS at WICHE. 1972.

1B CAMPUSICOLORADO is a version of CAMPUS VIII programmed for the CDC 6400. The main differences are thatCAMPUSICOLORADO has the PCS structure (developed by NCHEMS at WICHE) and has its awn formulae in its Revenue and CostingModules. Additional capabilities include the ability lo calculate he new space required for offices and the ability to maintain non-integerinventories of non-teaching staff. It has also additional dimensions especially in the number of cost centers.

Hussain and Mason

1' For a U.S. implementation. see William Lembus. "CAMPUS VII Helped a Small College Plan for the Future While it Had One" inCollege and University Business, Vol. 53, No. Nuvember, 1971. pp. 37-39. 58. In the same issue. are other articles on RRPM (pp. 32-33).SEARCH (p. 43), and CAMPUS (pp. '35-36).

1' For details on the logic and numerical examples illustrating the logic see K,M. Ilussaln. CAMPUS/COLORADO: Logic of the Model- A Sell Instructional Approach, Boulder. Colorado: Office of Institutional Research. V riiversity of Colorado. 1972, and K.M. Hussain, 1971,op cit., pp. 9-12, 70-92.

" See K.M. Hussain, 1971, op. cit., pp. 27-28.

1" See F.I. Jewett et at, The Feasibility of Analytical Models for Academic Planning: A Preliminary Analysis of Seven Quarters ofObservation of the "Induced Course Load Matrix," Los Angeles. Office of the Chancellor of California Stale College, 1970, and K,M. Hussainand N.S. Urquardt. The Induced Course Loud Matrix for Planning in o Community College: A Conceptual and Empirical Study. Hobbs,N.M : New Mexico Junior College, 1972.

" R.S. Johnson. NCHEMS Student Flow Model SFM-1A: An Introduction, Boulder, Colorado: NCHEMS at WICHE, 1972.

18 Huff. el al., op, cit,, pp. 41-54.

10 See G. Ziemer. et al., Cost Finding Principles and Procedures, Boulder. Colorado: NCHEMS at WICHE, 1971.

to For its implementation in conjunction with RRPM 1.6. see Huff et al., op. cit., pp. 21-36.

4' For details, see L.C. Romney, Information Exchange Procedures: Overview and General Approach, Boulder. Colorado: NCHEMS atWICHE, 1972.

a L.C. Romney, Faculty Activity Analysis: Overview and Major Issues Technical Report No. 24, Boulder, Colorado: NCHEMS atWICHE, 1972. pp. 81-82.

" Ibid.

21 For one research formulation of this, see T.W. fluent, Project Generalized University Model, Phase III, Final Report. Austin:Graduate School of Business, University of Texas. November, 1969.

4' For a sample. see W.W. Gulko and K.M. Hussain, A Resource Requirements Prediction Model (RRPM-1): An Introduction to theModel, Boulder, Colorado: NCHEMS at WICHE. 1971. p. '33.

46 See K.M. Hussain. 1971, op cit.. pp. 49-52, 65.

4' For details, see S.S. Mica and R.A. Walhaus, Outcomes of Higher Education - A Draft, Boulder. Colorado: NCHEMS at WICHES,August. 1972. and B. Lawrence, et al., (eds.). The Outputs of Higher Education: Their Identification, Measurement, and Evaluation, Boulder.Colorado: NCHEMS at WICHE. 1970.

Gary M. Andrew. op. cit., p. 11.

,0 D. Hopkins, "University Cost Simulation Models: A Critique." Berkeley, California: Office of Vice President - Planning andAnalysis, University of California, April 1969. p. 9.

144 D. Breneman. "The Stability of Faculty Input Coefficients in Linear Work Load Models of the University of California." Berkeley.California: Office of the Vice President - Planning and Analysis, University of California, February. 1969. Also reported by Hopkins. op. cit.,p. 13.

l'Culko and Hussain. op, cit., pp. 16-17.

For one approach using an LP model, see R.A. Walhaus. "A Linear Programming Formulation for Degree Output Planning," - ADiscussion Paper. Boulder, Colorado: NCHEMS at WICHE. 1971. See also A.M. Geoffrion et of.. Academic Deportment Management: AnApplication of on Interactive Multi-criterion Optimization Approach, Berkeley. California: Office of the Vice President of Planning October.1971. Paper. P-25. p.27.

" An example is Stanford University, with its major research component, as revealed in its pilot implementation of RRPM 1.2. See K.M.liussain and J. Marlin, op. cit.. pp. 75-82.

14 M. Alexander. "The Implementation of CAMPUS/COLORADO at the University of Colorado.- Forum of the Association forInstitutional Research. Vancouver. B.C.. May. 1973.

85

THE IMPLEMENTATION OF CAMPUS/COLORADO

Implementation of a simulation model means dif-ferent things to people at different levels and with dif-ferent roles in an institution. To the data analyst, imple-emntation is when the model "runs" and is validated. Tothe administrator, it is when the outputs of the model canbe used in a decision. To a department chairperson, it iswhen he or she can use it to justify additional money in thedepartment's budget.

Churchman and Schainblatt define the "problem ofimplementation" as "the problem of determining what acti-vities of the scientist and the manager are most appropriateto bring about an effective relationship between the two."'If one substitutes the words "analyst or planner" for scien-tist and "user" for manager and considers the number ofusers in a university setting, one has an inkling of the enor-mous and tong term task of trying to make a simulationmodel a useful tool in a university.

The purpose of this paper is to discuss how the Uni-

StructuralDefinition

StudentEnrollment

Madelyn D. Alexander, University of Colorado

versity of Colorado is implementing the CAMPUS ICOLORADO model at all levels within the institution. Thisimplementation began as a data problem, and this paperwill discuss this area in detail; but it soon grew to includethe projects necessary to carry CAMPUS to all its potentialusers including the State of Colorado.

About CAMPUSIt isn't possible to discuss the implementation as-

pects of CAMPUS without discussing, a bit, the nature ofCAMPUS.

Figure 1 shows the input to CAMPUS as being offour types:

1. structural definitions2. policy/planning factors3. student enrollments4. inventories

Structural definitions are used to "define" the insti-

Figure 1

.Input

0 w CAMPUS software

1.............j output reports

86

Inventories

ReferenceandAnalytical

tution to CAMPUS. These enumerate, for example, the costcenters (the organizational groupings of the model) andtheir relationships, the types of spaces, types of staff,program classification structure, budget functions, andother parts of the framework of the data to the model.

Planning factors are the expressions (variables, co-efficients, and relationships) which are used in projectingresource requirements. These can be input for each re-source type at each cost center. Policy factors are coeffi-cients which are input to modify resource requirements ac-cording to planned restrictions, expansions, or qualifi-cations.

Student enrollments are the projections which toadthe model. While CAMPUS has a student flow model built

Structural Definitions

Madelyn D. Alexander

in as an option, these can also be input for each year of asimulation from an external analysis.

Inventories is a generalized term for "what is." It in-cludes physical space inventory, staff inventory, other re-source inventory (non-personnel operating costs), and theinventory of activities offered. (An activity is one com-ponent of a course, i e. lecture, lab, recitation.) Theseinventories are used ir the model to show how estimatedresource requirements relate to existing or base yearfigures. It also allows for the requirements to be "adjusted"over compatible types of resources (a small class meeting ina larger room).

Figure 2 shows in more detail how the input relatesto the various modules (or parts) of CAMPUS.

Figure 2

Student EnrollmentProjections

ContactHour

GeneratorModule

Activities

= Input

Input Inventories

0 = CAMPUS Module

= Output Report

= Flow

Policy/Planning Factors

__e nventory

Support a nven ory'Indirect

ResourcesGeneratorModule

er esources nven ory

Multi-YearModule

CostingModule

87

CAMPUS/COLORADO

The magnitude of data elements required inCAMPUS is massive. The 1800 courses In one term' on theBoulder campus contained 2040 different instructional acti-vities in Fall, 1971. Forty-seven types of staff are definedand require identification of inventory, average salary, andplanning factor information in 100 cost centers. Buildingspace is defined by 70 types to be inventoried and plannedfor in 75 space control centers (specially designated costcenters). Other nonpersonnel resources are defined into 56types in 100 cost centers. There are 250 student programs (adefined student emphasis 'major] which leads to a degree,certificate, or some other stated objective) at C.V. atBoulder, which means that its induced course matrix is 250programs by 2000 activities by six student levels. One canquickly conclude that this is a nontrivial amount of in-formation. I'd like now to discuss how we approach thisdata collection process, understanding that it has to be doneevery year, without panic.

About the Data SystemsWherever possible, computerized data collection

conversion systems have been developed. It is my intent toserve four major purposes by this approach:

1. The amount of hand work necessary to generate theinput to CAMPUS Is reduced.

2, Intermediate reports are produced automatically.3. All input documents to CAMPUS where manual cod-

ing is also necessary are produced automatically withexisting information pre-coded.

4.A direct link is provided with operating managementdata systems.

The reduction of hand work carries significantimpact in a project the size of CAMPUS at Boulder. Thetime required to code the input would not only be long, itwould be prohibitive; and even a minor error rate in key-punching could carry significant alteration or the results.

During the conversion of operational data toCAMPUS input data, intermediate reports are producedwhich record this data as it is being converted. This inter-mediate reporting is needed for auditing, checking, andbasic information purposes. Even though CAMPUS is de-tailed, there is still substantial aggregation in some areas.For example, in the University's personnel system over 200actual job class codes are reduced into 47 types of staff forCAMPUS input. Data gathered is also "massaged" to pro-duce new kinds of data. Thus. CAMPUS forces us to look atthe data as we had never done in the past, and we neededan audit trail to "find our way."

When the conversion of data is complete, this data isrecorded on a report which becomes the input coding docu-ment to CAMPUS. Thus this data is pre-coded allowing forthe insertion manually of planning factor information.These reports produced by the computerized data collec-tion systems precisely parallel magnetic tape files of the

actual input to CAMPUS. Planning factors are added many-ally to the reports and these reports become the inputcoding sheets to CAMPUS, These manual additions arethen keyed on to the tape using a Viatron System 21.

The direct link with operating systems initiallyipro-vides a communications link with people working at theoperational level in two ways. First, it more clearly informsthe data system designers of the analytical needs of theplanning staff. This serves the long range goal that the"institutional analysis files are based on operating datasystems . . ."2 Secondly, this serves as the validation base.More will be said about validation later.

Examination of the operating systems revealed thatthey were supportive of most CAMPUS data requirementsfor the basic inventory systems, as well as the inducedcourse load matrix (1CLr1.

Figure 3 shows the flow of the data to the CAMPUSmodel. Aside from the policy I planning factors, the othermanually generated input represents less than 0.5% of theinput.

Some details of one of the automated inventory sys-tems is described below as an example. Each of thesystems is designed to take existing University operatingfiles and summarize the pertinent information to produceeither direct input to CAMPUS or partially completedreports which serve as coding documents with corres-ponding magnetic tape files.

Other Resources SystemOther resources are all non-salary operating ex-

penditures.The Other Resources Inventories are generated di-

rectly from the year-end Debit and Credit Register main-tained by the Finance Office which reflects the actual ex-penditures for the fiscal year.

Two translations are necessary:1. the account number translation to the appropri-

ate cost center;2. the object of expenditure code translation to the

appropriate other resource type.The expenditures are then summaryized for each

other resource type by cost center. For checking purposes,all salary expenditures recorded are also summarized foreach cost center end recorded in an intermediate report,Pre-coded input records to CAMPUS are generated.

There were, of course, many problems which besetthese best laid plans of mice and people. The first was thatthe elements within the data systems were not consistentlydefined. This is a problem which can only be solved by thelong range approach of common definition developmentand adoption. Al the University of Colorado, Administra-tive Data Systems is working in this area. It is our hope thatas these systems are redefined, the operation of data collec-tion for CAMPUS will become a normal spin-off from the

Space

Figure 3

Inventory

Madelyn D. Alexander

PrecodedInputForms

Personnel

Policy/PlanningFactors

Yes Change

OtherResources

($'s)

Key manually codedinformation to tapeby Viatron System 21

Activities

InputTape

CAMPUSMODEL

Students

Other Manually GeneratedInput:

Structural DefinitionsPolicy FactorsInflation FactorsEnrollment Protections

= tape files

ri = input

= process

= CAMPUS software

Keypunch

0= decision

89

CAMPUS/COLORADO

operating systems and be performed by the data processingand operating department staff; thus, totally dosing thegap.

Another data problem was finding historical dataconsistently maintained fot one year to provide a base case

for validation. Our first selection of 70/71 for a base yearhad to be changed to 71/72 when it was discovered that in70171 a census student course registration taco had not beenkept.

Figure 4

CAMPUS DATA COLLECTIONDOCUMENTATION LOG SHEET

Date

Author

Description (if pertinent):

For Ccmmand Code (if pertinent)

Sublect Area (check appropriate):

GeneralBudget

Cost CenterfacultyNonTeaching StaffSpiceOther ResourcesActivitiesV.rograrns

Enrollments

Items

Distribution List No

Abcdt Validation"Validation holds a special and important role in

simulation."3 While it is easy to believe in the "black box"of the model, unless the inputs can be related to familiardata, the user cannot and will not use the model because heor she will be unable to interpret projection data. Our basicchoice of working with operating data systems, even thoughmany problems were inherent in this approach, was tomake it possible to clearly define every path of dataaggregation andlor development. The intermediate reportsof the data gathering system have proved invaluable. Un-derotanding the data which drives CAMPUS ill toad inthe long range to an understanding of CAMPUS itself.

About DocumentationThere is a step which becomes the lifeblood of an

90

001.1

operation once it has gone beyond the first few monthsdocumentation, Dealing with a system as large asCAMPUS, it was immediately evident that as much aspossible should be documented. But how do you work thatmundane and lime - consuming operation into a dynamicsituation? By doing it along the way. By providing an easymechanism by which any staff member can quickly andwithout much bother record event3. The only real problemencountered at C.U. was reminding staff that documenta-tion had to be done! Figure 4 shows a form which was de-signed for the purpose of fast documentation free formin the documentation area, but structured so that only acheckmark made it "fileable" and "findable." This form,along with a similar one for recording telephone conversa-tions with SRG (Systems Research Group), has successfullycaptured the "history" of the project's decisions and analy

sis techniques.

About CommunicationThe activities of data collection and validation are,

of course, only the first part of implementing a model,Communicating with the users of a model is critical. "Inmost complex organizations including colleges Pnd uni-versities the executive officers, governing boP.rds, operat-ing managers, and policy committees are the ultimate usersof analytical models."4 Communicating with these groupshas r.onsumed much time and energy puticularly since theState of Colorado is also involved in our communicationprogram.

The use of transparencies provides a structured andformal format in presentation but allows for greatflexibility in the development of presentations. Five sets oftransparencies were developed by Visiting Professor K.M.Hussain of the University of New Mexi;o:

1. General information on CAMPUS.2. Introduction to CAMPUS / COLORADO3. How the model works.4, implementation of CAMPUS.5. Comparison between CAMPUS and RR -A.

A document which explains how CAMPUS workswas designed by Professor Hussain as a "stand alone" andincludes problem work sets for greater understanding. Inaddition, video -tapes were recorded on the general intro-duction to models (in addition to the five presentationsabove) and the comparison between CAMPUS and RRPM.

These presentations have been used with many audi-ences across the University community and the State,

Currently, however, we are working with four pilotdepartments on the Boulder Campus by extending to themdetailed, hands-on experience with the model for them,selves. This is being done jointly with the Committee on

Madelyn D. Alexander

Academic Planning (of the Faculty Council) in support ofan evaluation study of these four departments, By provid-ing the additional data available from CAMPUS to these de-partments in the open framework of faculty committeeproject, it is hoped that our communication will betied in a natural and positive way.

Our only concern at this point is that we should havecommunicated sooner with the faculty. "However, we be-lieve that it was necessary to have the computer softwareand the existing data basis in 'fair' condition before westarted discussions with the faculty."5

Some Facts and FiguresThe first year of implementation has Included not

only those items described in this paper, but also systemdevelopment of CAMPUS itself. Wo purchased CAMPUSVI with the understanding that it would be modified to in-dude such things as program costing. This was ac-complished along with the conversion of the programs torun on the CDC 6400. These development costs areincluded in the implementation costs presented below.

Direct Costs from March 1972 to Feb, 1973: $61,000Estimated Staff Costs from March 1972 to Feb. 1973:$40,000

In SummationWhy we chose to go the detailed Nutt; of CAMPUS

is adequately covered elsewhere. We did o this route andbelieve that the implementation design will permit thedetail to be handled automaticity while allowing forgreater flexibility in planning down to the departmentlevel, We also believe this approach to impIementation willlead to the slow absorption of CAMPUS / COLORADO intothe operational parts of the University where its impact willbe most useful.

Churchman. C.W, and A.H. Schainblall, The Researcher and the Manager: A Dialectic of Implementation," Management Science,February 1965.

4 Chaney, john F., Data Management and interrelated Data Systems for 'fisher Education, from the Seminar on the Advanced State-of.the-Art/The Sterling Institute, Washington, D.C., April 2426, 1969, WICHE, Oct. 19G9.

Van Horn. Richard L., "Validation of Simulation Results." Management Science, lanuary 1971.

Mason, Thomas R., The Purpose of Analytical Models: The Perspective of a Mode) User, University of Colorado, presented al SCUP1973 Spring Conference "Let's End the Confusion about Cost Simulation Models!", Washington, D.C., to be published.

Andrew, Gary M., CAMPUS at Colorado, University of Colorado, Boulder Colorado, presented at SCUP 1973 Spring Conference "Let'sEnd the Confusion about Cost-Simulation Models!", Washington. D.C., to be published,

91

SIMULATING THE UNIVERSITY;AN EVALUATION OF EXPERIENCES WITH THE CAMPUS, HELP/PLANTRAN AND SEARCH SYSTEMS

Recent years have seen much activity in simulationmodeling for research and planning, and several organi-zations have developed simulation models designed es-pecially for use by administrators of higher education.Although the power and sophistication of this tool havebeen proven in areas of business and government, manyquestions remain to be answered about the effectiveness ofthese models as an aid to administrators of highereducation.

The ultimate goal of this application of the .chniqueof simulation has been to enable colleges and universitiesto make more rational decisions about the use of thou ownresources and the direction of her development. Rourkeand Brooks have found that the extent to which this expec-tation has been fulfilled in higher education is as yet farfrom clear. While the literature reveals many articles re-ferring to the use of simulation as something of Et panacea,other writers have expressed doubts concerning the suit-ability of its application at a level of complexity com-parable to that of a,:iministering a university.

As the number of colleges and universities usingsimulation models continues to rise, it becomes increas-ingly important to examine objectively the reactions to thisnew science. This paper provides an assessment of theutilimion of three computer simulation models now beingapplied to administrative problems in selected institutionsof higher education.

Method of ProcedureThe problem of evaluating the use of simulation

systems was approached through an examination of the ex-periences reported by colleges and universities that hadimplemented and were utilizing simulation as an adminis-trative tool.

A sampling technique was used to achieve thefollowing objectives: (1) the selection of the Institutionswhich represented a variety of orgar izational structures,types of methods of utilizing simulation models, and (2) the

92

Jerome F. Wartgow, Governors State University

selection of computer simulation systems that were repre-sentative of those most widely used by institutions ofhigher education.

The criteria applied were as follows: (1) the instittlion must have had irnplementcd the model at least oneyear prior to the dale of the study, (2) the simulation modelbeing used must have been readily available and adaptableto utilization in any institution of similar size and organiza-tional structure, mid (3) the institution must be an insti-tution of higher education in the United States.

When the criteria were applied to the institutions, itwas readily determined that only three simulation modelswould be involved in the study, These models were identi-fied as (1) CAMPUS. developed and implemented bySystems Research Group, Toronto, Canada, (2) HELP IPLANTRAN, developed and implemented by Midwest Re-search Institute, Kansas City, Missouri, and (3) SEARCH,devloped and implemented by Peat, Marwick, Mitchell &Co., New York, New York,

The final sample consisted of (1) a total of eightinstitutions, (2) each of the three simulation systems beingutilized by at least two institutions, and (3) a variety ofinstitutional SIZO3 and types, including two large privateuniversities, is. two-year college, a private women's college,a state university system, and three small private liberalarts colleges. The sample is not random, but rather a repro-sentation sample of institutional types and the simulationmodels they were utilizing.

The study incorporated accepted descriptive-surveyresearch procedures in the collection and analysis of data.Personal semi-structured interviews were conducted ateach of the participating institutions and experiences withthe respective simulation models were analyzed under theheadings of: (a) factors influencing the purchase and imple-mentation of the model, (2) experiences during imple-mentation, (3) means and methods of utilization, and (4)the extent to which the simulation model had achieved itsobjectives.

FindingsWhile experiences during the implementation and

utilization of computer simulation systems varied frontcollege to college. ceain common factors were identifiedwhich appeared to contribute to the extent of successfulutilization. Based on these findings a number of conclu-sions and recommendations were drawn. the more signifi-cant of which aro summarized below.

Two primary factors were identified whichinfluenced the decision to purchase and implement a com-puter simulation model; (a) the effort of an individual onthe university staff who had a personal interest in newtechniques of management. and (b) a recognized need byuniversity personnel for a tool to assist in answering "whatif" types of questions.

Two inferences may develop out of these findings.First, it might be inferred that the technique of simulationis so now that its existence is not common knowledgeamong administrators of higher education. As a result, theuse of simulation is promoted by that type of individualwho is aware of and interested in innovative managerialtechniques generally.

A second inference might be that the relevance ofsimulation to administering higher education will not begenerally recognized until the planning process on thecampus has progressed to a point where administrators areforced to decide between alternatives and are aware of theneed for information concerning the implications of choos-ing one alternative rather than another, At that time simu-lation will be recognized as a useful tool in assisting deci-sion-making.

Once purchased and implemented, simulationmodels were utilized more extensively in those institutionswhich purchased the system to meet a recognized need.than in those institutions which purchased the systemprimarily because of the recommendations of an adminis-trative "innovator."

No discernible pattern was evident concerning thedecision to purchase one of the models in preference to an-other, although it was clear that much of the discussioncentered on choosing a model of appropriate complexity.

The confusion over the desired complexity or simpli-city for a given simulation model reinforced a generallimitation of simulation that can he identified in the litera-ture, That is, that in order to represent the system accurately, there is a tendency to develop more complicated models.As models become more complicated they become less easyto unaorstand and thereby defeat the purpose of construct-ing a midel (i.e,, simplification of the real system to facili-tate understanding },

Experience Peported by Ins'ttutions D'iring Implementa-tion of the Model.

The time to make the system operational was signifi-

Jerome F. Wartgow

cantly underestimated in each implementation. Severalproblems were identified as contributing to this condition,The first and most significant was that considerable modifi-cation was required to several of the models prior to utili-zation. This factor may have far-reaching implications forthe use of simulation in higher education. It might be in-ferred that "simulation packages" specifically designed forone institution are not readily adaptable to other insti-tutions, This inference would support arguments advancedby some administrators concerning the "uniqueness" ofhigher education and the resultant inapplicability of sci-entific management techniques.

However, when the findings leading to this con-clusion are examined carefully, it appears that another in-ference might be more valid, That is, that the immediatemodification in models arises from the fact that the sophis-tication of the user has not developed to a point where hocan identify appropriate and Inappropriate users of themodel. A review of the literature strongly indicated that thevalue of any simulation system is largely dependent uponthe ability of the user to determine situations in which it isappropriate. Extensive modification of the system mightimply that simulation Wil3 being applied to problem forwhich it was not appropriate.

A second factor that was identified as contributing tothe length of time involved in implementation is the inexperience of the personnel responsible for using thesystem, The assumption that a person with no prior experi-ence with simulation or computers can operate a simulationsystem would appear to he invalid. The findings of thestudy indicate that inexperienced persons may implementthe models eventually, but that if simulation is to be con-ducted efficiently, the user must have had some prior ex-perience in the use of simulation or computers, or both.

A misunderstanding of what the term "implementa-tion" implies may have been a third factor which contri-buted to the discrepancy between actual and estimated im-plementation time, Quite possibly, implementation has adifferent meaning for the firms installing the simulationmodel than it does for the institutions that are utilizingmodel,

Institutions that relied primarily or, university per-sonnel during implementation experienced more difficultythan institutions that utilized the services of the firms thatdeveloped the models. The least difficulty during the im-plementation was reported. by those institutions that contracted the entire implementation to outside personnel,Generally, the problems encountered during implementa-tion were in the areas of data collection and computer tech-nology.

A lack of wide and active participation by universitypersonnel during initial stages of implementation appearedto influence the extent of future utilization of the model.

From this finding it may be inferred that partici-

93

CAMPUS, HELP/PLANTRAN AND SEARCH

pation leads to a fuller understanding and eventual accepLance of the model. This would support the argument thatresistance to the utilization o "Scientific Managementtechniques" stems mostly from men who misunderstandthe nature of modern administration. Further, in those in-stances when participation In the development of themodel was limited, and use was also limited, one mightinfer that the model was comparatively unsuccessful as aresult of misunderstanding on the part of those who werenot involved.

In-service sessions and in-service materials influence the extent of utilization of simulation models. Inaddition to increasing the number of sessions and person-nel involved, there was seen to be a greater need for improving the content of some of the in-service sessions. Con-sideration and discussion of the "human element" as it re-lated to the use of simulation was seen to by as important tosuccessful model utilization as technical considerations.

Means and Methods of UtilizationSimulation was most extensively utilized when a

formal planning process was in operation at the universityprior to its implementation. The case reports suggested thatimplementation of a simulation model prior to properpreparation tended to complicate rather than clarify its rolein the overall planning process. In this regcrd, simulationwas not a substitute for planning but rather a tool to beused to supplement the planning process.

It seems clear that simulation will have relevanceand applicability to higher education only when a lesssophisticated system of planning has already been usedsuccessfully, and when the acknowledged need for furtherand more detailed analysis leads to an understanding of thespecific circumstances in which this technique is appro-priate.

The amount of confidence placed In the accuracy ofsimulation results (by those persons in a position to makedecisions based on these results) was found to be a func-tion of bread participation by the institution's personnel inthe development of the mode/ and confidence in the individual conducting the simulation. Greater confidenceamong the users of simulation results was noted whenthere was wide and active participation in the developmentof the assumptions and f3rmulas used in the model. Addi-tionally, confidence held by university personnel in theability of the administrator responsible for conductingsimulation was found to be correlated to confidence in thesimulation itself.

The accuracy of the base data in the initial use of themodel tended to influence the exteN of future utilization aswell as the degree of use confidence in future simulations.University personnel often view the initial use of simu-lation with cautious skepticism. The careful collection ofaccurate base data will enhance the potential for successful

94

utilization and consequently help overcome the skepticism.Based on tho findings of the study, it was not

possible to draw any conclusions concerning the most ap-propriate institutional office to be responsible for primaryutilization of the model.

However, one observation may be relevant. General-ly speaking, it appeared that the personal and political in-thience of the indivicual having responsibility for the simu-lation model was a more significant factor in successfulutilization of the model than the position which heoccupied.

Insufficient time to devote to planning was identi-fied as a major impairment to the utilization of simulationmodels.

ConclusionsGenerally speaking, it is suggested that those persons

who are responsible fur development and/or use of uni-versity simulation models evaluate the model in terms offour basic criteria.

PERFORMANCE lbw effective is the model ingetting the answers I want?

UTILITY - How useful is the simulations system?How often will it be used? Is it flexible enough to acceptmajor changes In organizational structure? How manypeople can make use of d?

TIME - What is the time required for in,Itallation?How much time is required for collecting base data neces-sary to operate the system? What is the time required to re-trieve information?

COST - Is the value of the information worth thecost of implementation? Will it save money in terms of timeand personnel? Do we really need one at current costs?

If the criteria are applied to each of the availablemodels, much frustration will ba. eliminated. Additionally,if persons who are responsible for building new modelsutilize these criteria, the potential of simulation for ad-ministering higher education will Increase significantly.Overall, experiences to date have indicated that the timeand expense involved with simulation models have notbeen justified in terms of the extent of their utilization.However, this conclusion must be considered in thecontext of the following qualifications: (1) an importantbenefit of the utilization of these models is that attention offocused on long-range planning, (2) the models have theygreatest potential of becoming a valuable and appropriatetool in institutions which are in a process of change. and(3) the value of computer simulation models in highereducation is dependent upon the ability of the user torecognize situations in which this tool is needed and appro-priate,

When there was an existing and recognized need fora tool to assist in determining the impact of alternativecourses of action, the model was seen as a relevant and

applicable tool. On the other hand. when models wereapplied to a problem prior to a need for the tool heitig evi-dent, major difficulties were encountered,

In the opinion of administrators who used simula-lion, efficiency of utilization will increase as the user be-comes more familiar with the advantages and limitations ofthe system. Additionally, there was an expressed "feeling"that, just as during implementation of any new adminis-trative technique, a certain time factor is necessary to workout problems, gain the confidence of the staff, and over-come resistance to change.

As those conditions are met and proper preparationsmade, computer simulation models wilt have the potentialof becoming a valuable administrative aid. With the passingof time and the satisfaction of certain other stipulationswhich have been identified above, the potential should beachieved. Al that time, the use of computer simulationmodels in the administration of higher education will pro-vide valuable assistance in the task of more efficiently allo-cating institutional research.

RecommendationsIn an attempt to capsulate my opinions concerning

problems and prospects of computer simulation models, Ihave formulated a series of recommendations. Empiricallyspeaking, I would say that consideration of these recom-mendations by institutions planning to proceed withdevelopment and/or implementation of a computer simu-lation system will help eliminate some of the problems andbrighten some of the prospects.

A specific need and a high level commitment toplanning should be generally evident in the institutionprior to implementation of a computer simulation system.What happens on campus prior to implementation :s just asimoortant in terms of successful utilization as whathappens following implementation.

An institution sho.ild carefully select a model orsysten, that is best suited to the unique needs of the indi-vidual elllege or university. Care must be taken to insurethat the model is not too simple to represent the institutionadequately or too complicated to be easily understood.

An institution should take care lest it underestimatethe time necessary to make the system operational follow-ing the decision to purchase. Experiences reported indicatethat the time necessary to operationalizc the system isgenerally significantly longer than the initial estimate.

An institution or installing agent should define andclarify what is meant by the terms "installing," "imple-menting." and "operationalizing" the model. Experiencehas shown that these terms often have a different meaningfor the firms installing the model than for the institutionthat is using the model.

An institution should encourage wise and activeparticipation and involvement with the model from the

Jerome F. Warigow

outset. Lack of participation by the institution's personnelin the development of assumptions and formulas to be usedin the model is a strong predictor of unsuccessful utill-zetion.

An institution should employ a person to physicallyoperate the system who has prior experience with models,computers, or both. The findings of the study indicate that,contrary to prevalent assumptions, a person with no priorexperience in these areas cannot efficiently operate thesystem.

An institution should employ the services of one ofthe professional firms to implement the system. Experiencehas shown that this may be less costly than attempting thetask solely with college or university personnel.

An institution or installing agent should thoroughlydiscuss the rationale for any initial major technical or con-ceptual modifications in the models. It may be possible thatthe model is not appropriate for the institution, or morelikely, that the problem being approached is not appro-priate for the model

An institution should be prepared to evaluate in-service sessions and in-service materials to be sure that thecontent as well as the number of sessions is adequate tomeet the needs of the institution. In-service sessions shoulddeal with "human elements" which may cause problemsduring utilization, as well as technical details necessary tooperate the system,

An institution should provide adequate time for theadministrator who is responsible for utilizing the model toperform this function, Experience indicates that purchasinga system and giving it to an administrator as a "spare-time"activity is a poor investment.

An institution may establish confidence in the re-sults of the system by placing the model in the office of anadministrator whose iudgment is respected and who has anappropriate level of personal and political influence andprestige within the institution. Although this may causeproblems if the individual leaves the institution, ex-periences indicate that the confidence that university per-sonnel have in future simulations is positively correlated to'ihe confidence obtained as a result of the initial utilization.

An institution should take extreme cue in the collec-tion of base data to be used in the model, Do not attempt toconduct any studies with the system until the staff hascomplete confidence in the validity and accuracy of thebase data. If the institution is not experiencing majorchange of one form' or another, it should carefully weighthe value of a computer simulation model against the timeand expense involved in its purchase and implementation,

An institution should only use the system in areasfor which it is more appropriate than other techniques. Theability to determine when simulation is, and is not appro-priate, is the primary determinant of its value to theinstitution,

eN1 0.1

95

RESOURCE PREDICTION MODELS:A DO-IT-YOURSELF APPROACH FOR INSTITUTIONAL RESEARCHERS

No well-informed administrator questions thevirtuoso attributes of the well-known simulation models.Yet their potential utility is questionable in many institutions. The anamoly is easily explained. The modelstend to be internally driven especially by enrollment,But the crucial issues for today's colleges and universitiestend to he external: attitudes toward higher education,national economic fluctuations, demographic trends. thevagueries of politics. These are issues which require solidresearch and continuous monitoring if their influence upona particular institution is to be understood, Such researchand monitoring often are inadequate or wholly kicking.One is surprised to find in the modern world collegeofficers and governing boards who perfer envelope backs toresearch studies, but they are plentiful in smaller instituLions and not unheard of in major universities, Howeverreluctant top management may be, colleges and universitiesarc now being forced by circumstance to pursue willo'-thewisp, modern management techniques. Their initial forayssometimes are into simulation models. At times, andperhaps frequently, the results must reaffirm our favoriteaphorism, "garbage in, garbage out,'

The developers of elaborate simulation models stressthat the utility of the models is constrained by the data fedinto them, and that data on the future are especially elusive,NCHEMS, for example, in its "A nlueprint for RRPM 1.6Application" (WICHE, February 1973;, warns:

The fifth step of the implementation process in-volves the difficult task of developing alternativesets of specific planning decisions that use thestatus quo reports as a point of departure. Manyindividuals including faculty, students, lay boards,and administrators will probably plat, a part inthis process of developing alternative plans. Thetechnician must be available to the decisionmakers to translate their decisions into therequired RRPM input data formats and to carry outthe sixth step, which is making the RRPM projec-

Webster C. Cash, Fisk University

lion runs for each alternative set of planningdecisions.

Others besides NCHEMS have already pointed effectivelyto the need for good historical data and soundly conceivedfuture data if simulation models are not to be misleading,even dangerous, as decisionmaking tools,

How are the dangers reduced or avoided? One wayis for the institutional researcher to take the lead indeveloping simulation models which ore easily understoodby officers and governing boards, and which are tailored toa particular institution's circumstances and policy issuesa do-it-yourself approach,

In 1972 we developed at Fisk a do-it-yourself modesfor Fisk's financial projections even though we knew thatin 1973 we would begin operating the !:.:HEMS InducedCourse Load Matrix, the RRPM 1.6, and the CostFindingPrinciples. The model immediately proved valuable iniosolving policy issues. It is still valuable as a complementto the three NCHEMS systems, which are now operating.

The need for projecting the operating budget into thefuture was perceived toward the end of the innumerablelong, dreary, depressing meetings of administrators andsenior faculty in which we refined our knowledge of theUniversity's academic and fiscal problems, We wereconcerned that a trend of financial deterioration beginningin the latter 1960's might not be reversed quickly enough,and 14e were dissatisfied with our lone five-year financialprojection. We needed to explore alternative fiscal pathsand we chose to do so by developing a model which wouldsimulate Fisk's operating budget five years into the future.

The Model and the Computer ProgramExamine Exhibit One, The items on the left are the

major parameters which concerned us as we discussed theUniversity's finances. They are, I believe, typical of themajor fiscal concerns in most small institutions. Each itemin Exhibit One is subject to change either by policy

revision or by the Influence of national conditions. Hencethese items became variables in the model. They areIndependent variables that is, each can vary independently or other variables, and none of them aredetermined by the others. A change in any independentvariable affects one or more items in the operating budgetas well as the budget totals, The dependent variables arethe lino items of the budget.

Fisk's budget officer then sketched the model. Itsessence is to specify precisely what dollar changes willoccur in the operating budget as a result of a change in anyindependent variable. Drawing from his knowledge of ourrecent financial experience, ty2 completed that task in a fewhours, The following weekend he transformed the modelinto a computer program.

The elements of the model are:The Fisk budget for 1972.73 (major categories only)supplies the base amounts from which the amounts forfuture years are calculated.

(2) Each major category in the operating budget is pro.jetted through the five-year period endin3 with 1977-78.

In each trial projection, quantities are set for each ofthe following 20 independent variables whateverquantities the user of the computer program chooseswhich become inputs into the computer program. SeeExhibit One for an example,a. Enrollmentb. Student - faculty ratio.c. Net square feel used in activities budgeted against

educational and general items.d. The minimum number of square feet of education

and general space permited for each student; thispermits a calculation of enrollment capacity.

v. Tuition and fees per student.f. Room and hoard charges per student.g. The dollar income from federal programs each

year, distributed between:(1) Ninety percent to the revenue item "spon-

sored research and programs,-(2) Ten percent to the revenue item "in-

direct costs."h. Expenditures of federal program funds, divided

between:(1) Percent supporting Fisk's regular, on-

going programs of instruction and ad-ministration (all of which are E &G items).

(2) Percent supporting institutes, research,and other activities which are not funda-mental to Fisk's regular, on-going academicand administrative programs.

i. Endowment value.j. Percent of endowment value used as endowment

support.

(1)

(3)

Webster C. Cash

k. Private gifts itnd grants allocated to the operatingbudget excluding amounts budgeted under thenew Ford Development Program, and gifts andgrants for capital programs.

1, Income from the Ford Foundation's DevelopmentProgram allocation to the operating budget.

m. Federal subsidies to higher education Institutions,as estimated on the basis of Higher Education Actof 1972.

n. An inflation factor affecting all expenditure items.u. Percent increase per student for instructional in-

provement over and above increased costs result.ing from inflation,Average faculty salaries.Library expenditures calculated as a percent oftotal FAG expenditures,

r, Expenditures on financial aid to students, expressed as a percent of income from tuition andfees (this is called "full-aid equivalent" because IIis equivalent to the percent of students onfinancial aid if every financial aid package con-sists of the full cost of tuition, fees, room andboard, and if no smaller packages were awarded).

s. An efficiency factor for administrative expenditures allowing real costs (that is, dollar expen-ditures measured in constant dollars) to beadjusted dowro., ard as administrative resourcesare, used more efficiently,

1. Number of students residing in dormitories.(4) A set of simple equations instructing the computer how

to use these inputs to calculate the "correct" amountfor each major item of the operating budget in eachyear through 1977.78,The resulting calculations for each trial projection,shown on a computer printout as Fisk's operatinghial.get (major categories only) each year from 1972.73through 1977.78.

Si me 40 different projections each with a dif-ferent set of input data were run on the computer. Onlyone of these projections is reproduced here in the threeexhibits.

p.

q.

(5)

The Base Projection: Current PoliciesTo understand the starting point for the projections,

the reader should examine the exhibits. This "base projec-tion" is a test of the iciasibilily of Visk's present policies ifthey were 'o remain unchanged over a five-year period. Thedata and msumptions for the base projection were selectedaccording to three criteria:

First, that all decisionmaking until 1978 will he inaccordance with Fisk policy as established in the sutnmerof 1972;

Second, that in matters not governed by policy, theassumptions used as inputs in the computer program

97

DOIT-YOURSELF APPROACH

Exhibit 1

TRIAL PROJECTIONS OF OPERATION BUDGETPROJECTION NO.1

PURPOSE: To lest the feasibility of Casting policies Combined with con lye estimates of future trends.

ludgc.fivt.Te atIncrease

Input of variable data 197141 197.3.74 1974.75 197346 197677 1977.78 or Decease

Nonrtionelary:I. Average enrollment 1500 1575 1650 171$ 1715 1725 + 15%2, No. of Eft faculty 110 1C8 108 108 108 108 .1%3. Ratio of students to faculty 13.6,1 14.6:1 15.3:1 16:1 16:1 16:1 + 10%4. Net square feel to E&G uses 198.000 200,000 100,000 230,000 240,000 250,000 + 26%

Revenues:S. Tuition & fees per student $1750 $1855 $1969 31085 $2110 $2140 +34%6. Room S board fees per sludent $7375 $3289 $1354 $1410 $1490 S1S6S + 18%7. Federal sponsored programs

a. % to indirect costs$2.0 A.i.

10%$2.2 mi.

10%31.3 ml.

10%$2,4 mi,

10%

32.5 mi.10%

$2.6 mi.10%

+ 30%constant

b. % to instilulional support 35% 35% 35% 35% 35% 35% constant8. Endowment value $7.4 ml. $7.4 mi. $7.4 mi. $7.4 ml. 37.4 mi. $7.4 mi. 0%9. Ford Development Program $700,000 $530,000 $405,000 $305,000 9155,000 $155,000 .78%

10. Total private gins & giants egfudingford Development Program $1,295,096 S1,419,000 $1,180,000 31,161,000 $1,139,000 $1,357,000 + 5%

11. Federal ,ubsidies to private colleges 0 0 0 0 0 0 0%

Espenditures:11. Inflallon factor for all espenditure

items 5% 5% 5% 5% 5% 5% +20%13. % increase per student for Instructional

Improvement 2% 2% 21. 2% 2% 2% + 10%14. Average faculty salaries $14,150 $15,140 $16,200 $17,314 $18,548 $19,846 +40%IS. Total faculty salaries $1,556,500 $1,639,120 $1,749,600 51,872,071 91.003,184 $1,143,368 + 38%16. % of student body awarded financial aid

(full.ald equivalent) 30% 30% 30% 30% 30% 301 0%

'The bs,dgeted amounts for 1972-73 were preliminary; they later were revised subslantially.

Exhibit 2

FISK UNIVERSITYTRIAL PROJECTIONS OF OPERATING BUDGET

PROJECTION NO.1

PURPOSE: To leg the feasibility of existing policies combined with conservative estimates of future trends.

budget'Five.ib_arIncrease

REVENUE 1972.73 1973.74 1974.75 1975.76 1976.77 1977.78 or Decrease

Education and General:Tuition & Fees $2,593,779 31,773341 33,080,137 53,416,793 33.611,637 $1,1334,674 + 40%Sponsored Research & Programs 1,803,530 1,980,000 2,070,000 2,160,000 2,150,000 i,340,000 +30%Endowment Support 445,200 471,340 472,340 472,340 472,340 472,340 +6%Gifts S Grants (excluding Ford Dev. Pro.) 1,295,096 1,419,000 1,180,000 1,162,000 1,139,000 1,357,000 + 5%Ford foi,ndation Development Program 700,000 030,000 405,000 300,000 155,000 155,000 7111Federal I. Tenn.Institutional Subsidies 0 0 0 0 0 0 0%Indirect federal Support 187,612 210,000 230,000 240,000 250000 260,000 + 39%Other 334,746 351,483 369,057 387,510 406,685 417,230 +18%

Subtotal Education S General $7,359,959 $7,748,367 $7,806,535 58,141,644 $8,294,864 $8,846,246 +20.1%

Ausgivy Enterprises:Dining & Dormitories $1,492,637 $1,086,974 $1,667,249 S1,753,699 $1,840,149 $1,932,774 + 1S%bookstore, Real Islet, Miscellaneous 268,420 287,918 316,710 347,662 355,045 383,297 +43%Athletics S IMelo 41,950 19,407 32,436 35,606 37,386 39,356 4%

Subtotal Augtiary Enterprises $1,803,007 31,914,3111 $2,016,397 $3,136,968 51.142,3112 $1,355,3111 +30.6%

Financial Aid:Subtotal Federal Grant, (MG, Work

Study, M'scellaneous) 9 852,230 S 93S,38" 51,033.541 $1,114,147 11,191,2.'4 $1,250,038 + 46.8%

TOTAL REVENUE 310,013,1% $10,592,331 $10.856,473 511,415,159 SI 1,728,720 $11,432,412 +24.3%

NOM Suidotals and totals may vary %lightly from the /pp/refill, correct amounts because of rounding.

'The F,dgeted amounts for 1971-73 vete preliminary; they later were revised substantially.

98

Webster C. Cash

Exhibit 3

FISK UNIVERSITYTRIAL PROJECTIONS OF OPERATING BUDGET

PROJECTION NO.1

PURPOSE: To lest the feasibility of 'Killing policies Combined with con lye estimates of future trends.

ludgetfiveVearIncrease

IXPENDITURTS 1971.73 1971.74 1974-75 1973.76 1976-77 1977.78 or Decrease

Education and General:Instruction 6 Organized Activities S 2,115,050 S 2,203,591 $ 2,353,888 $ 2,515,022 $ 2,693,589 5 1,864,814 +36%Sponsored Research 6 Programs 1,371,961 1,287,000 1,345,500 1,404,000 1,441300 1,521,000 + 11%Library 371,691 313,013 140,967 365,618 392,261 417,235 + 02%Student Services 616,082 646,886 679,230 711,192 784,511 823,737 + 34%General Administration 411,450 495,021 519,773 545,762 600,338 630,355 + 34%CatvelopmentAlumni.Unlversity Relations 169,380 198,848 208,791 219,231 230,192 241,702 +28%General Institution & Stall gene fits 611,512 642,087 674,191 707,901 778,611 817,626 +34%Physical Facilities 624,840 663,818 697,009 841,639 921,143 1,008,594 +61%

Subtotal Education 6 General 5 6,372,966 5 6,460,248 S 6,819,351 S 7,112,367 5 7,865,228 $8,145,104 +30.1%

/military Enterprises:Dining & OW171110405 $1,467,176 51566,974 51,667,249 5 1,753,699 51,840,149 $1,932,774 +12%lookstore, ittat Estate, Miscellaneous 261,151 287,911 )16,710 341,662 365,045 383.297 + 47%Athletics, Radio 106,985 117,950 124,146 142,425 149,457 157,024 + 47%

Subtotal Auxiliary interpriser

financial Aid:

5 1,835,312 5 1,992,844 5 2,113,706 5 2,243.781 5 2,354,742 $ 2.41 ,096 +34.1%

Subtotal khoiarships & Other Grants $1,106.871 S 1,481650 5 1,640,425 5 1,813,837 5 1,914,750 5 2,029,837 + 18%

Contingency - S 250,000 5 250,000 5 250,000 5 250,000 5 253,000 -TOTAL EXPENDITURES 5 9.915,151 $10,186,762 $10,823,984 511.619,991 $12,384,720 $13,089,087 + 32 0%

Surplus of revenues over espendoures S 100,045 5 405,568 5 32.488 - - - -Additional revenues required - - - 5 204,832 5 656,000 5 636,625

NOTE: Subtotals and totals may 7iiy slightly from the apparently correct amounts because of rounding,

'The budgeted amounts for 1172.71 were preliminary: they later were revised substantially.

should reflect our estimate of the conditions and trendswhich will prevail during the projeltion period; and

Third, that where the most probable conditions ortrends" are seen as a range of data in every year rather thanas a single series of data, the assumptions should reflectconservative estimates - usually Interpreted as changescausing income estimates to be near the lower side of theprobable range, and expenditure estimates to be near theupper side of the probable range.

The resulting total revenue (Exhibit Two, last row)rose very modestly indeed - only by 28.1 percent in fiveyears. This result deviated abruptly from the experience ofthe recent golden years. We looked for errors. None werefound. There could be only one interpretation: that if Fiskpolicies should persist unchanged through 1977.78, and ifcur estimates of non-policy treads were not unnecessarilyconservative, then the trend of large increases in revenuefrom all sources, seen continuously over more than adecade, had ended in 1972-73.

The increase in total expenditures (Exhibit Three,third from last row) was accounted for largely by the in-flation factor. A five percent increase when compoundedfive years becomes 28 percent, and total expenditures roseonly 38,2 percent.

The last two rows of Exhibit Three gave us theanswers which we were seeking. The surplus of revenuesover expenditures would have a short life, if any. The ad-ditional revenues required - a euphemism for the word"deficit'" would mount rapidly. The rise in the deficit wasmore dramatic than we had expected.

Fisk's position here differs little from that of mostleading private colleges and universities which haveanalyzed thoroughly their financial positions. They arecurrently operating approximately in balance and they caneven predict : occasional surplus. But their fiscal stabilityis in the words of Mr, Cheit, fragile.

Our model had given us quantitative insights intothe trends. Because of the considerable research that under-girded the projections, we felt that they were about asaccurate as any projections can be, Something had tochange - and quickly. The question was: What? And sowe continued to run the program, altering the 20 variables.

Uses of the DoItYourself ModelOur goal then became to discover a combination of

policies which would keep the budget balanced withouteducational retrenchment. The model allowed us to seek asolution through an iterative process. After innumerable

99

DO-IT-YOURSELF' APPRAOCH

computer runs, each of ten minutes duration, we chose acombination of policy variables, which we believed was anoptimum combination, We did so in the light of severalinstitutional studies concerned with major variables in themodel, The variables so chosen became our recommen-dations on fees, enrollment, faculty, size, space. gifts andgrants, and allocations to major expenditure categories,Many of the recommendations were incorporated in thebudge.' for 1973.74, and some became University policy to1978.

Comparison with NCHEMS SystemsThe NC}IEMS ICLM, the RRPM 1.6 and the Cost-

Finding Principles are even newer to us at Fisk than is thedo-it-yourself model, Our initial reaction to them is that theseveral models are fine complements to one another.Together they appear to give beautiful results for decision-making.

The RRPM 1.6 supplies cost detail in abundancesuch as the direct costs of programs. measures of facultyproductivity, and requirements for non-faculty personnel. Itlacks the indirect costs of educational programs and thecosts of their research components. From our point of view,however, it has only (..1- e major deficiency: it is notconcerned with income.

The NCHEMS Cost-Finding Principles, whichconverts expenditure data into program costs, yields resultswhich are especially well suited to decision-making inyears of fiscal stringency. Because of the considerabledetail yielded by this program and by the RRPM 1.6, weview both of them as models or systems for the microanalysts of costs.

13y contrast, the do-it-yourself model involves onlymacroanalysis a broad-brush treatment of budget projec-tions but it analyzes income as. well as costs. Its advant-ages, we have found, are:(1) That it was written to give specific answers to specific

issues of a specific university;(2) That it forces faculty, administrators, and trustees to

face major issues of policy and to revise major assump-tions while not bothering them with details which theymight refuse to study;

(3) That its results are readily perceived and understoodby officers and trustees who are used to thinking interms of major budget categories;

(4) That it supplies an analysis of income;(5) That it has a quick turn-around time and is convenient

100

(7)

to operate;That projections are not from unit costs they are

non-linear; allowing economies of scale and otherfluctuations; andThat it was inexpensive to develop and it operates atvirtually no cost. (It was developed at a cost of lessthan $1,000 in the time of staff and about $50 for thecost of operating the computer.)

The do-it-yourself model's disadvantages appear to be;(1) That it lacks the detail required to resolve numerous

issues and policy matters (the NCHEMS systemssupply much of this needed detail on the expenditureside);

(2) 'That when its results are misleading, this may not be,iorent immediately; and

That when its results are prima facie erroneous. ittends to conceal the source A of errors,

(3)

The Role of Institutional ResearchThe role of institutional research in planning at Fisk

is to pursue studies which supply information andestimates about the critical planning parameters, to assistcolleagues in focusing squarely upon the main issues of theday and of the future, and to interwine institutional studiesand planning documents st, that the two functions of insti-tutional research and planning often are indistinguishable.We believe that the University benefits from this in-tertwining of functions, and we are beginning to amass evi-dence which will demonstrate these benefits conclusively.If the do -it- yourself model does indeed have the utilitywhich we believe it has, and if the NCHEMS systems proveas useful in planning as we expect them to, these advant-ages should be attributed in some measure to theundergirding supplied by institutional research to the plan-ning processes and to the intertwining of the two functions.Time will tell.

We have no major concerns about the NCHEMSsystems, which are well-received on campus. We arepleased with them and with the do-it-yourself model, Weknow of course that such systems and models can bedeveloped and operated with input data from envelopebacks. But they deserve better treatment. As decision-making tools, they aro, not sufficient unto themselves.Healthy doses of institutional research are the missingingredients, and they can best be supplied if planning andinstitutional research are intertwined or, better yet, if theycare merged.

PLANNING MANAGEMENT SYSTEMS IN HIGHER EDUCATIONSWILL THEY REALLY MAKE A SIGNIFICANT CONTRIBUTION TOIMPROVING INTERNAL DECISIONS?

Much attention has been given to the development ofManagement Information Systems and budgetingmethodologies, including Programming, Planning, andBudgeting (PPD) and cost finding principles. Much moneyhas been spent on developmental programs. Even moremay be spent on implementation.

It has been said that within a few years institutionswill be required to submit data in NCHEMS form to qualifyfor any governmental funding. There is real questionwhether data of the NCHEMS magnitude and complexityare really needed by the government and whether theirbroad category statistical compilations that come out lateare worth the total cost of data collection and preparation.Even more important is the question, Will the underlyingManagement Information System provide much usefulassistance in internal decision-making? Please note that I

distinguish between reality and theoretical possibility.Those of us who have been involved, either directly

or indirectly, in the development of large Managementinformation Systems may have become overly enthusiastic.We are like some corporate product-managers. We havedone some market research in needs and wants, but wehave given insufficient attention to ways in which thecollege may want to apply the product and to organizationalenvironments in which the product a must be used. Wehave brought forth a tool designed for a flexible hand withan opposable thumb. But the organizations for which it isintended have only a paw.

There are various reasons PlanningManogementSystems may have more benefits claimed for them thanthey can deliver. These reasons might be classified roughlyas first, technical, and second, organizational, sociological.and psychological.

Consider the technical reasons from the standpoint,of the three main parts of a Planning-Management System

the data file, a package of standard reports generatedfrom lime to time, and a set of models and analyticalpackages, such as a resource prediction model and cost-finding programs.

The data file: For economic reasons the data file willnot be, and should not be, complete. Completeness meansthat one has everything one could conceivably need, This

Peggy Heim, &mime 11 University

often means coding things different ways indeed, tens ofdifferent ways. They may also have to be broken down intocomponent parts not just one set of component parts, butmany different sets, for even the same user may want theanalysis prepared different ways. And each part willprobably require several pieces of supplementary informs-lion for analytical purposes. Updating so largo a data basebecomes cumbersome and costly. The data user will there-fore have to accept something loss than what he wants orthe additional costs of customtalloring analyses andreports. Either situation will reduce the usefulness ofPlanning-Management Systems.

Standard reports: Wo will discover that regardless ofhow many ways wo count things, data users will claim theyaro not meaningful because something should have beenexcluded or something else included. On the other hand, ifwe produce all the variations that might conceivably boneeded, users will bo awash with reports. If overwhelmed,potential users will not refer to our products either.

Models and analytical programs: There are obviouslyquestions on the usefulness of models:

First, modeling technology is best adapted to rela-tively static conditions, but we actually face a period ofsignificant economic and social change. Mayor variables arechanging rapidly prices, wages and salaries, workingconditions, modes of instruction, course patterns, and off.campus study to name a few. Results will be no better thanthe model itself, Even the more complex models andsystems may yield disappointing results, This occursbecause formalized models lack flexibility and versatility,Simulation models or cost finding models are acceptableonly if the players accept the rules of the game. There isconsiderable question whether the potential data users --the various administrators, department chairmen, andfaculty members want to accept the same rules. There isalso doubt whether they even want to play the same game.

A second reason to question the usefulness of Plan.ning-management Systems is that they understandably giveconsiderable emphasis to cost-finding programs. These areaverage costs and average inputs. But the significant factorsfrom the standpoint of internal decision making are theincremental inputs, costs, and outputs. They may bear little

101

PLANN1NGMANAOEMENT SYSTEMS

relationship to average costs or outputs, Moreover, the bestwe seem to do with measuring outputs is to count some-thing, such as the degrees conferred by major, scores ongraduate record exams, percentage entering graduate orprofessional school, and so on crude devices formeasuring quality.

A third criticism of simulation mode's and costfinding programs is that they are too mechanistic. We focusour attention on the wrong factors and the wrong objec-tives. This occurs because we fail to define the problemproperly. We are more likely to isolate a factor and thendefine that factor as a problem, Take faculty tenure as anexample. The bask problem is not really one of the propor-tion tenured but of providing flexibility, vitality, and appropriateness of program, Tenure is only one part of a muchbroader problem. Its control therefore provides only onefacet of a solut,on. 3lnce individuals focus more easily onsimple mechanistic factors than on complex problems, anysystem which facilitates a simplistic approach mayinadvertently lead us down the primrose path. In over-looking the basic problem, we miss more effectivesolutions,

Even if we did not encounter these technicalporbiems, there is another problem, namely the applicationof PlanningManagement Systems in a complex socialorganization. Since you are familiar with the variousrequisites for the successful use of Planning-ManagementSystems, there are only two points I shall make.

First, the appropriate college administrators must bewilling to come to grips with painful decisions, particularlythe President, These decisions usually involve people, Buttoo often decisions are not made. The institution drifts. Orbecause they involve people, Ihe decisions are made thewrong way.

The second requisite is the distribution of respon-sibility, theoretical authority, and practical power. Theintroduction of Planning-Management Systems would beone thing in the classical organizational pattern (thepyre-nid) where authority is concentrated at the top, anddelegated, along with responsibilities, to lower levels of ad-ministration and where job retention and salaries dependupon job performance. But our institutions of higher educa-tion are organized on a collegial principle carried to thepoint whore many, if not most, institutions haveadministration by committee. The committees have mixedconstituencies, depending upon the institution and theparticular committee. In most institutions faculty continueto exercise strong, if not dominant, influence within thecommittee, and thus within the institution. It is importantto note that most of the Committee members the faculty

are employed to perform other functions, like teachingand research not to manage the institution. Their jobs do

102

not depend on how well they mako administrativedr ,4;ons. There are also many sociological reasons whythey might not be expected to make hard decisions and touse management data in making them. But even if we araable to educate them, the committee membership changes.Each year we start the educational process all over again. Inview of these problems I suspect it will take considerabletime before educational institutions begin to derive greatbenefit from Planning-Management Systems

We now come to the hard question: What can the In-stitutional planner do? What should he do?

(1) Do not oversell Planning-Management Systems,Tell them how it might help them; but tell tilem how Itmight not.

(2) Do your own market research with key administrators. Attempt to identify problems which disturbthem problems which seem important to theadministrator and which they believe should be tackled,Then try to assist them with data, If the administratorwants something calculated three more ways when you arealready calculating it two ways, calculate it the five ways.The ones they are likely to use are the other three ways.

(3) Attempt to identify problems of critical longrunimportance to the institution. What typo of action you canhelp bring about once you identify the problems will varyfrom situation to situation. Three pints should be kept inmind.

(a) Progress comes slowly, The tortoise is morelikely to win than the hare.

(b) Have empathy for the higher level admini-strator. Remember, he is standing in front of abigger fan! Or more fans!

(c) Even if you identify successfully problemswhich are critical to the institution, there may bevery good reasons why nothing is done about them

at least, in the manner in which you would seeit done. An administrator cannot take on too manyissues at one time. Unfortunately, some admini-strators want to take on as few as possible.

(4) Attempt to offer alternative solutions orstrategies. This is particularly important when dealing withpeople-problems. Although the alternatives may seem lessdesirable than the basic solutions, action which goes partway may be better than none,

(5) Look down the road. Attempt to anticipate theturns and the potholes. The important thing is to anticipateand plan for the probable consequences before the derisionis made, Next most important get lead time in order toexpand the institution's alternatives.

Finally, remember that the important things will notrequire a complex PlanningManagement System, but theywill require both analytical and diplomatic skiil.

4,,

THE COMPARABILITY QUESTION:POTENTIAL USES AND MISUSES OF DATA

Higher education has long enjoyed a measure offreedom from external accountability which is envied bythose who are outside the system and probably not fully ap-preciated by those within it. This priviledged position isnow being penetrated on all sides by a variety of calls foraccountability, usually in terms of specific data by which itis assumed that effectiveness and efficiency of operationcan be determined.

The fiscal crisis now facing most public and privateeducational institutions has not been the solo reason forthis cry for accountability, though it has probably had thegreatest impact. Students are more aware of what and howthey %are taught and are often heard calling for greaterrelevance, Many state legislatures are looking more closelyat public higher education in their states In terms of itscontent and conduct as well as its cost, At least oneGovernor is calling for faculty to increase teaching loadsand the general question of the continuation of the tenuresystem as we know it today is being more frequentlydiscussed, A major concern of all these groups isimprovement of the educational experience through greaterstudent-faculty contact, smaller classes, more acces-sibility to distinguished professors and accomplishmentof this at the lowest possible cost,

The common approach to the analysis and resolutionof these issues has been tc require more and more data bywhich complex systems can be quantified, summarized,and compared. This is understandable at a time when theconsumer is increasingly concerned with getting thegreatest value for his dollar. What is not fully understood isthe extent to which such data, if not compiled and analyzedwith the greatest care, can lead to totally inaccurate or in-appropriate conclusions.

The purpose of this paper is to discuss, throughactual examples, the difficulties encountered when at-tempting to make routine comparisons of data and tosuggest how data comparison can be handled meaningfully.In so doing. I will also point out how such data can bepotentially misunderstood and misused. It is not my intentto discount in any way the need for accountability itself. Itis entirely appropriate that higher education should be ac-countable, albeit in different ways, both to those whobenefit from its services and to those who finance its

Adrian H. Harris, University of California, Los Angeles!

operatinns, There is also no question but that we will allbenefit from such an effort, provided that it is conducted ina responsible, informed, and conscientious fashion,

The National Center for Higher Education Manage-ment Systems (NCHEMS) has now appeared on the sceneto lead the field in the development of products intended toaid in the reporting process. Unfortunately, there is atemptation to use the results of these efforts as ends tothemselves. Such a cookbook approach can more often leadto inappropriate conclusions than to informed judgments.This is particularly true when the result is a single number,such as a unit cost figure, which once obtained can beeasily misused unless there is a full understanding of whatis being presented and how it can and should be used,

Among the products developed by NCHEMS: aro theProgram Classification Structure which allows institutionsto divide their programs into standard disciplinary units;cost finding principles which attempt to set standards forthe determination of unit costs; RRPM (Resource Require-ment Prediction Model) which has been mistakenlyassumed by some people to be a model to develop unitcosts for comparative purposes instead of its realmission as an institutional planning model for internal use;and the Information Exchange Procedures project which isdesigned to attack the problem defined in its title, Suchmodels are valuable to an institution's internalmanagement, decision making, and evaluation. The dangeris in assuming that such techniques can produce readilycomparable results when applied to different institutions,

ProblemsThus far I have dealt with the question of data com-

parability in a general sense. I would now like to offersome specific examples of the problems involved in theselection, definition, and analysis of data to be exchangedor compared. The examples are primarily drawn from theexperience of a six-campus data exhange effort in which Iwas involved for over a two-year period. The six campuseswere all large state supported institutions having broadlybased programs of undergraduate and graduate instructionand research. In pursuing the initial goal of data exchange,we uncovered many unforeseen problems.

The first difficulties in data exchange are defini-

103

COMPARABILITY QUESTION

Hanel. In our own effort wo spent countless hours on thetask of understanding the individual use of terms and thentrying to produce meaningful decision rules on the classi-fication of data,

A commonly used data element is the full -hmoequivalent student (student FTE). In ma group, all in-stitutions but one determined undergraduate student FTEby dividing student credit hours by 15; the other used adivisor of 15.5. Of course, greater variation will occurbetween institutions which have different requirements forgraduation or different definitions of "normal progress." Atthe graduate level the problem of student FTE is vastlymore complicated because of the differences in graduateprograms and the manner in which graduate education isconducted and credited. For example, one institution in ourexchange did not attach any credit value to doctoral work.As a result of these problems, we finally agreed to let eachinstitution use its own method of calculating student FTEand to accompany data with a chart defining the methodused, Obviously, anyone using these data must fullyfamiliarize himself with these definitions and understandtheir differences before drawing conclusions.

A related problem is the need to determine whetherstudent FIE' should be aggregated by student level or bycourse level. Within our group, there was considerabledisagreement on this matter. Some felt that course level isthe significant determinent of workload; others felt thathigher level students create more workload for facultyregardless of the level of course being taken. An addedproblem was the fact that one campus did not identifycourse level and thus could not provide data in this form.In either case the choice can has significant effects ondata, particularly when micro programs are beingcompared, and thus the use to which the data is to be putmust be considered in reaching a decision.

Another definitional problem centered around thedifferences in University calendars and the time period forwhich data are to be exchanged. Among our six-campusgroup it was found that some used a quarter system, bothwith and without a summer quarter; some used a semestersystem; and one used a trimester system covering the fullyear. Because of these differences, we agreed to excludesummer enrollments from the academic year. Annual datamust also be defined as covering either a fiscal year,academic year, or calendar year While student informationis usually easily separated by academic years, financial in-formation cannot always be treated in this way. Forexample, the campus mentioned above which used a year-round trimester system encountered enormous difficultiesin attmpting to exclude its summer component to make its"annual" data comparable with the other institutions,

The results of our deliberations over these and amyriad of other problems of the same nature wererepresented in twelve pages of definitions covering the data

104

elements to bo exchanged and spelling out institutional dif-ferences in detail. Tho complexity of meaningful dataexchange is well illustrated by this document, which maybe obtained by writing to me,

By comparison, the attend edition of NCHEMS'Data Element Dictionary: Student devotes 125 pages to 73data elements; their comments give little idea of theproblems we encountered and, of course, this is only one offive such dictionaries.

Selecting Comparable DepartmentsOur next stop in the exchange of data was

identification of a series of academic units for comparison.Our initial intention was to select departments that were asnearly Identical on all six campuses as possible, rec-ognizing at the outset that programmatic differences maycreate difficulties In interpreting the data. Thus it wasunderstood that, while most universities are organized intoschools, colleges, and clop?r monis, not every field Isrepresented at every institution Furthermore, disciplines,departments or programs having the same name may notinclude the same subject matter, This is an Importantproblem which is not readily apparent to the outside userof data. It is obvious that physics and English are different;it is not so obvious that two departments called "English"may vary on different campuses. In our own group it wasfound that some English departments included rhetoricand/or speech and some did not; thus English was rejectedfor comparison, Similarly, romance languages could not beused because in some cases French was included and inothers it was not, At UCLA, for certain historical reasons,our Department of Linguistics Includes instruction inseveral exotic languages not covered elsewhere. Somecampuses identify a separte Department of ComputerSciences; at UCLA, instruction in computer sciences isIncluded within engineering and cannot be separated. Onthe other hand, the same program may be called bydifferent names at different institutions; environmentalscience on one campus might be ecology on another. Theremay also be differences between the way graduate and un-dergraduate programs are organized in some fields, Aftersifting through the known differences of this kind we wereonly able to select seven departments for inclusion in theinitial data exchange. Later, when analysis of the datauncovered differences even among these departments, itwas agreed to provide departmental profiles outlining theprogram and requirements for each. It was evident thateven with careful preliminary screening for similarity,these seven departments also displayed a mixture ofobjectives, resources, and organization as they wererepresented on the various campuses.

As we continued our efforts, it became increasinglyclear that meaningful data exchange required a full under-standing of programmatic elements. Not only was it

necessary to understand the individual programs beingcompared, but the institution within which they were setand the entire system of higher education as well. institu-tional priorities, and commitments, outside pressures forchange, and Oven national interests may all have an Impactwhich is reflected in the data for a particular program at aparticular point In time, The individuals involved in thiseffort were all high-level administrators in their respectiveuniversities involved in institutional research andplanning. In spite of this, we oi!en found that further in-vestigation was needed to supplement our knowledge ofprogrammatic and organizational differences, Imagine,then, the difficulties that would be encountered by oven themost conscient)us outsider attempting to analyze our data.

ResultsThe final step in this initial effort at data exchange

was convarison and aitalysis of the results. Many of thedifferences which were observed were explained by closerexamination of the programs being compared. Only a fewof the many possible examples will be discussed here.

One measure compared was an unweightedstudent/faculty ratio. In the case of one department, the sixinstitutions reported values of 11, 15. 15, 16, 17 and 34, In-vestigation of the differences here showed that the 34 wasthe result of misinterpretation of definitions, an occurancefor which one must always allow. On the other hand, thesmall number was associated with a new program; a newprogram might easily be expected to have a smallerstudent/faculty ratio due to pre-staffing, in order to coverthe field, before enrollments have reached their peak.

The sttldentlfaculty ratios In art departments wererelatively consistent for all institutions except one, whichhad half the ratio of the others. In this case the departmentshowing the lower value placed a relatively greateremphasis on art practice than on art history. Thus, classeswere required to be taught in small studio sections andmore faculty were needed to staff them, This also had theeffect of reducing the budgeted support per faculty memberto less than half the level observed in the other art depart-ments. A somewhat higher student/faculty ratio at anotherinstitution, resulted from the high service load associatedwith a "general undergraduate art requirement."

In some cases, a department showed a high numberof graduate students per senior faculty. While on thesurface this appeared to be a commendable use ofresources, it apparently resulted from the production ofvery few doctorates over a number of years. If the presentgroup of students received their degrees within a shortperiod of time, the data would be sharply reversed.

Similar situations occurred when support funds perbudgeted faculty were compared. For physics, the resultswere especially interesting; the six campuses reported thefollowing figures for this depa:tment; $12,000, $10,000,

Adrian H. Harris

$5.000. $3,000. $14,000, and $11,000, The two campuseshaving extremely low levels of support par budgetedfaculty in this case were associated with large extramurallyfunded programs. On the other hand, the campus reporting$10,000 per budgeted faculty was also known to have alarge extramural program including a well-equipped radia-tion laboratory, yet it eld not show the low support figurethat might have been expected, Our examination alsoshowed that this institution's program was ranked firstamong all physics departments in the country.

At one Institution the support level in law was fortypercent above the others. In this case the departmentalbudget included prevision for a law library; at the otherinstitutions this was part of a separte library budget.

Psychology departments having a strong clinicalemphasis characteristically had support levels as much astwice as high as those emphasizing the social sciences,Here four of the campuses reported support per budgetedfaculty at $4,000; the other two reported $7.000 and $8,000.In the first case of higher support the departmentcharacterized itself as "having more of a laboratory sciencethan social science emphasis." The other was stated to be"associated with a large clinical program." In both cases,the higher figures were easily explained by the costsassociated with the support of laboratories and specialfacilities, a reflection of programmatic differences.

Programmatic differences can affect data oven whenprograms are aggregated by discipline. Another examplecompares disciplinary groups on two campuses of the sameuniversity. For the biological sciences on one campus thesupport per FIE faculty was recently reported as $16,000 atthe other it was $12,000; however, (he second campusincludes a vigorous health science program which isbudgeted separately. In the same report the physicalsciences on the first campus showed a support figure of$13,000, compared with $17,000 at the second. Here, thefirst campus benefited from a large externally-fundedresearch facility.

ImplicationsThe examples could go on indefinitely; the implications,however, are clear. While many legitimate differences willbe observed when data are compared among institutions,many other differences are created by conditions notdirectly related to the data element being compared. Thiswas recognized by the RRPM1 Task Force of WICHEwhen they stated In a resolution published on March 9,1971;

Institutions of higher education differ widelyamong themselves. They have differentapproaches to teaching, different degrees, differentrequirements for the same degree, different coursemixes within a single institution to satisfy therequirements for the same degree, different course

105

COMPARiIIILITY QUESTION

contents, different methods of awarding andcomputing credit hours, different support acti-vities, different stadent/faculty ratios, differentgoals, and other differences too numerous toenumerate .5

An understanding of these differences, and consequentlyany meaningful comparison of data, can only beaccomplished by persons intimately acquainted with thedata collection procedures, the programs, the budgetingpatterns, the organization, the objectives, and ot'ierimportant characteristics of the institutions involved.

One might think that the most meaningfulcomparison of data would be made in an unchangedenvironment at a single institution where various mea-sures are compared for the same units over a period oftime. However, such conditions never exist. Even in theleast dynamic setting, change occurs and one cannot drawsimple conclusions collected over time. On the other hand,given an understanding of the system and of thesignificance of differences between units and of changeswhich may be occuring, data comparison can be animportant decision making tool for an institution. Inter-institutional comparisons are vastly more complicated, andmust be undertaken only with the greatest understandingand care.

Universities will continue to be faced with amultitude of requests for cost analyses and other data. Solong as resources for higher education remain scarce, thoseresponsible for their allocation will continue to look forinformation to help them with their difficult decisions, It islikely that institutions wilt also be required to demonstratethat they are using their resources in the most efficient andeffective manner possible.

Unfortunately, many people feel more secure whentheir decisions are based on quantitative rather thanqualitative information. However, where higher educationIs concerned, considerations of quality are (or should be) ofgreat importance. While qualitative considerations areusually understood to imply judgments that are moresubjective than objective, there is no reason for them to beuninformed or capricious. In addition, it must also berecognized that even when the end retult is a number, anessentially subjective decision may have been made in theselection of the method for producing that number.

Numbers are not magic. They do not tell the whole storyany more than a picture of one side of the moon showswhat is on the other side, Numbers can, however, be useful;if they are properly understood and interpreted they can becombined with other information to givo a reasonablycomplete picture.

We cannot escape the fact that data will be collectedand comparisons will be made, What we must seek to avoidis standardization and regimentation around formats anddefinitions designed by those far removed from the systemin an attempt to assure "comparable data," Ultimately, wewill have to live with resource allocations and otherdecisions that may be based on such efforts. However, it isessential that higher education not be reduced to the leastcommon denominator,

We have found from practical experience that anumber of conditions are absolutely essential to anymeaningful exchange or comparison of data in highereducation. At the outset, the effort must be designed andundertaken with the full participation of knowledgeableindividuals representing all parties involved. Sufficientparticulars must be provided to allow rational accomplish-ment of the task. Specifically, they must seek to:

Fully agree to the purpose of the comparison andthe use of which the data will be put;

Insure that the data to be collected aro consistentwith the expressed purpose;

Define all terms explicitly and include detaileddefinitions with any data disseminated;

Provide profiles of each program to be compared,including information on all aspects ofprogram which may affect the data; and

Permit the providers of the data to review theresults so that errors of fact or interpretationmay be avoided,

Obviously, this is an ideal and not every conditioncan be met in every case even when undertaken with thebest of intentions. Hopefully, a better understanding of theproblems and concerns discussed here will help toestablish a meaningful environment in which thoseinvolved in higher education and those to whom they areaccountable can work together responsibly andresponsively.

t Acknowledgment is due Miss Dee Cuenod, Miss Gertie Ewing, Mrs. Jeffrey Gilbert, Dr. Wayne Smith. Miss Corrine Verhuist(members of the UCLA Planning °like Staff and other members of our data exchange group for significant contributions which lead to theproduction of this paper.

Publications describing each of those items are available from the National Center for Higher Education Management Systems atWiCHE. P.O. Drawer P, Boulder. Colorado t0302.

-1 The other data element dictionaries lire for course, facilities, finance, and staff,

106

Adrian Harris

4 Kenneth D. Kane and Charles I. Andersen, A Bating of Grachate Programs, American Council on Education, IN70. This institutionwas ranked first for Quality of Graduate Faculty and fifth for Effectiveness ut Doctoral Program and in each ease ranked higher than any of theother programs compared.

s Derived from 1972.73 University of California Budget,

Appendix A of "Report of RRPM implementation at UCLA," UCLA Planning Office, September 1971,

107

AN INFORMATION BASE FOR BLACK HIGHER EDUCATION

Institutions categorized by the U.S. Office of Edu-cation as developing, by the mass media as Black colleges,and by themselves as historically Black colleges and uni-versities, have for many years carried out an extremelyvaluable societal funtion. This function is the education ofthe Black youth of America, Such a function was the as.signment of a young American educational society notquite certain of its destiny and unsure of its place in worldhistory. Consistent with the impetuousness of youth, theconclusion which this young American educational societydrew as it related to the education of its Black popula-tion was that Blacks required a separate educationalexperience. Thus, the evolution of the historically Blackcolleges and universities.

Historical records, most of which are buried in state,federal, and some college archives, reveal much of theprogress of these institutions. In recent years, however, asignificant effort has been launched by the presidents ofsome 113 of these institutions in the establishment of a databank to contain facts and figures on their institutions. Itwas their wish that such a data bank, once constructed,would not merely function as a repository of historicaldata on themselves, but also as: (1) a model for individualinstitutions to utilize in fashioning an institutional researchdata base and (2) a central data source for educational his-torians and researchers studying the historically Black col-leges and universities. The effort referenced is the Tech-nical Assistance Consortium to Improve College Services(TACTICS), and an agency called the ManagementInformation Systems Directorate (MIS) of the Institute forServices to Education, Inc. (ISE) in Washington, D.C.Funding for the MIS directorate of ISE was provided by theU.S. Office of Education. Generally stated, TACTICS is toprovide the historically Black colleges and universities witha range of technical services which historically have notbeen within their individual financial ranges. Training inboth workshop and seminar form are available in the areasof Admissions, Financial Aid, Recruitment, CooperativeAcademic and Administrative Planning. Media Tech-nology, and Information Systems. The MIS directorate wasassigned the responsibility of developing a data bank ofsome specificity and long-range usefulness which shallprovide the participating TACTICS institutions with the

108

James Weich,,Institute for Services to Education

capabilities cited above. The statements which follow shallput forth what MIS/1SE construed as being implicit in thisassignment, and what to dale, has been developed as aresult of such an assignment. 'The relevance of this de-velopment to the topic, "An Information Base for BlackHigher Education," should bec.Jme apparent as the discus-sion of theso developments unfolds.

MIS/ISE Developed Information SystemMIS/ISE drawing on its staffs' experience, the ex

perience of the executive and senior program associates ofthe company, advisory groups consisting of college pros!.dents, technical personnel, and senior administrators andfaculty from both within and outside of the TACTICS insti-tutions, designed and developed data base approach whichin their collective judgment would be initially useful for acollege's institutional research.

It was in the collective opinion of the various ad.visories that the educational / administrative environmentof a developing college or university would require,Initially, a set of data elements, which, when properly mas-saged and interpreted, would yeild as accurate a description as possible of an Institution's internal status. Thesedata elements, to yield such a description, should,therefore, encompass the following institutional environ-ments:

StudentsFaculty and StaffFinancial (Income & Expenditure)Physical FacilitiesCourse /ProgramInstitutional CharacteristicsMIS, with the concurrence of the various advisory

groups, recognized that there would be three basicproblems in collecting data of this nature:A. Though many of the terms in higher education were

familiar to persons in institutions of higher education,consistent definitions of these terms were not avail-able.

B. Collecting data of such depth and detail from institulions which are unaccustomed and/or not presentlystaffed to provide same, would require in most in-stances, major procedural and/or policy alterations on

the part of the institution(s).C. Initiating and implementing an IR office for insti-

tutions, the largest percent of which have no electronicdata processing (EDP) capability, would requireunique staff capabilities on the part of both MIS andthe participating institutions.

These three basic problems were approached andsolutions sought.

Problems Associated with Data Base DevelopmentA Data Elements Dictionary was developed by the

directorate along with members of the technical advisorygroup of MIS. This dictionary provides definitions, descrip-tions. and examples of some four hundred terms or wordsesoteric to the field of higher education. The rationale forthe development of this dictionary was to assist individualsin the institutions participating in the MIS/ISE program insurpassing the translation problems associated with in-consistent definitiont of words, terms, and phrases relatedto higher education. In almost every case, new words orphrases were not invited, but instead, the most fr )quentlyapplied and existing definitions to educational for s wereutilized. The U.S. Office of Education, the Americrn Asso-ciation of University Professors, and the educationaltaxonomies in conjunction with USOE's HEGIS (I-PgherEducation General Information Survey) definitions weresuggested and utilized. An additional function of the dic-tionarywith some expansionwould be to function as abasis for the establishment of a "lingua franca" for educa-tional researchers. Three dictionaries were mailed to eachof the 113 TACTICS institutions participating in the MISprogram during its first year of operation. The dictionaryapproach, therefore, appeared to satisfy the first basicproblem of inconsistency in educational terminology,

Problems associated with the collection of detaileddata on and in developing institutions o.cre approached intwo ways, First, agencies which historically have querieddeveloping institutions were identified, and with the unani-mous support of the presidents and chancellors of theTACTICS institutions were authorized to release all perti-nent data related to their institutions, to MIS/ISE. Once thisdata was collected and analyzed. it was determined that itscontent was inadequate to form the appropriate basis uponwhich could be constructed an IR function. Questionaireswere, therefore, designed and developed to supplement thecollected data and were mailed to the various colleges anduniversities. The two design requisites for these ques-tionaires. bearing in mind the problem of the institution notbeing geared to respond in any great amount of detail, wasto: (1) develop a survey instrument which would be usefulto the responding institution as a research profile of itself;and (2) provide an instrument which, when completelyfilled out, could be utilized as a guide for responding toquestions on other surveys.

James Welch

An additional problem associated with data collec-tion was revealed through the analysis of the agency-collected data. Inconsistency in defintion of terminology fromone year to the next was the most prominent problem. Themost consistent base was data gleaned from the HEC1S(Higher Education General Information Survey) reports ad-ministered by USOS /Natlonal Center for EducationalStatistics. MIS, together with its advisors, determined thatthe most lucid data base could be formed by coupling theHEGIS data with data collected through the separately administered MIS/ISE: survey instruments. The MIS/1SE databank, therefore, contains data pertinent to the areas out-lined earlier In this sectiln on 113 developing institutionsdating from the present (spring, 1973) back to Fall of 1965.

It is, with this base providing the input, that MIS/ISEhas developed an Online QUERY computer system usefulto institutions desirous of implementing an automatedinformation base for institutional research and manage-ment information purpcses.

The Evolution of a SystemIn the opinion of the MIS directorate, it is inconceiv-

able to believe that just because a set of procedures hasbeen developed which may improve an institution'sresearch operation, that that institution is able to performall of the required policy and procedural alterations neces-sary to accommodate it. To have, therefore, a system avail-able to colleges desirous of implementing an automated IR,tailored to their individual needs, and one which couldmore closely conform to a developing institution'seconomic resources, was the undergirding approach to thesolution of the third basic problem conceptualized by thedirectorate initiating and implementing an IR office forinstitutions, the largest percent of which have no electronicdata processing.

As of January 31, 1972, there were operativo 42 com-puter configurations in 38 of the 113 TACTICS institutions.This statistic indicates that about 70% of the institutionsparticipating in TACTICS had no computer capability. Todevelop an automated system for IR purposes in view ofsuch facts required additional consideration, Such con-siderations included: (1) the overall IR and informationmanagement objectives of the various institutions: (2) theoperating budget of the institution's IR office; and (3) theinstitutions aggregate staff capability. Based on thesefundamental considerations. the remote-access, computertime-shared approach. was construed as being the mostappropriate for the largest cross-section of the TACTICSInstitutions. Given this decision, a model system was de-veloped and is now operative for use in the MIS Wash-ington, D.C. office. The model system is presently referredto as QUERY.

The design and documented procedures which aninstitution may follow for the establishment of an insti-

109

INFORMATION BASE

tutionally-based QUERY System approach, therefore,assumes access to a computer facility via some remoteaccess device teletype or console keyboard. Someadvantages of an automated system i think would be inorder here because non-technical people often concatenatecomputers with magic, Tho principal advantages of ther.ystem being automated are; (1) a large data base could homaintained; (2) the manipulation of the base would notrequire as many man hours as it would if it were manuallymassaged nor would the number and kinds of manip-ulations be resrained duo to their complexity; and (3) theconstruction and testing of various educationally valideconomic models would not be restrained either by time orcomplexity of models duo to a proliferation of variables, ineach instance, one quickly notices that the advantages ofthe automated over a manual system is the amount of timeconserved and the elimination of human error due to theredundancy in the data manipulations, Whether an insti-tution requires repetitive data manipulations or time conservation should be entirely their decision. Thus, thesystem being developed by MIS shall be available for thoseInstitutions requesting it,

You should be asking yourselves, "How do these de-velopments contribute to the subject, An Information Basefor Black Higher Education?" My response is this:

First. The development of the Education Data Ele-ments Dictionary provides educators as well as researcherswith a basic definitive reference for educational terms andphrases.

Second. The design and development of a survey,

110

which if properly understood by tho institution, wouldfuncticn as an institutional profile report internally and asan external report on the institution to MIS. Such a docu-ment is presently being utilized as input into a data basefor the TACTICS colleges and universities.

Third, Tho collection of agency data (I.e ISE,Higher Education General Infermation rArvey, etc) addsanother historical dimensior to the information on theTACTICS institutions which should be extremely Interest-ing when paralted with data gleaned directly from thecolleges.

Fourth. Given the volume of analysis necessary tointerpret the collected data the QUERY system was de-signed as a data organization and retrieval system whichwould enable a researcher to get to the data required for thegeneration of institutionally useful reports. Such reportswould be productive in assisting institutions to plot theirprogress over any particular time period.

iso

In Summary

You should now begin to see how these develop-ments fit together, Seven years of data on 113 institutions isin one location, A system is being developed so that a re-searcher does not necessarily have to be in that location toretrieve specific kinds of information. Tho files in thesystem shall be so structured that if one institution wishesto develop such a system for itself it may use the developedsystem as a model, or if a group of institutions would wantto create an IR network, a system would be available,

FINANCING UNDER CONDITIONS OF DECLINING FLIAISILITY

James R, Montgomery, John Janney, Gerald W. McLaughlin,and David Slog, Virginia Polytechnic Institute

When an institution becomes aware that disparityexists between its financial expenditures and resources, itmust, if it is to remain viable, make preparations for func-tioning in the midst of such imbalances while simul-taneously initiating programs designed to rectify thesituation. II is lust such a situation in which increasing

FINANCIAL INPUT

FEDERALSTATEPRIVATECOMMERCIAL

STUDENTSFACULTYSTAFFADM IN !STRATI ON

numbers of institutions of higher education now find them-selves. There are, furthermore, strong indications that thistrend will continuo for the foreseeable future.

As an aid to discussion of this problem, a simplifiedsystem diagram is presented below. Included are theprimary concepts involved. The problems facing adminis-

Figure

SYSTFM DIAGRAM FOR HIGHER EDUCATION

CAPITAL ASSETSSERVICESSALARIES, WAGES,BENEFITS

UNIVERSITYANDENVIRONMENT

INPUT INTERACTION

KNOWLEDGE & IDEAS ---->

MORE KNOWLEDGEABLEPEOPLE

MORE WELL-ROUNDEDPEOPLE

PUBLIC SERVICE

RESEARCH FINDINGS

CULTURAL SERVICE

BY PRODUCTS

COMMUNITY RELATIONSSOCIAL ACTIVITIESACADEMIC FAILURESDISMISSALSSTUDENT PROTEST

111

FINANCING

trators in higher education are perhaps best treated in termsof three distinguishable, if overlapping, areas of concern:financial (imitations. enrollments, and social responsibility,With respect to the first consideration, there is little argu-men' that institutions of higher education, both public andprivate, have simultaneously been confronted by spiralingcosts and resta:ted funding. For publicly controlled insti-tutions in the United States, the current expenditures perfull-time-equivalent student (in terms of constant dollarsfor 1969-70) increased from $1456 in 1959-60 to $1896 in1969-70.1 This increase, after correction for inflation duringthe decade, strongly supports the claims of administratorsthat costs have mushroomed,

Sellers' MarkelThe decade of the 60's saw a sellets market for

college faculty and staff which was largely 11 e product ofsharp enrollment increases and heavy spending on re-search. In addition, while opening fall enrollm9nt for allstudents in all institutions of higher education n.ore thandoubled over the decade, other factors including inflation,rising faculty salaries. increased expenditures for studentaid, theft, and destruction continued to boost costs.' 3

A comparison of the large universities of theseventies with institutions of twenty-five to thirty years agoindicates faculty spend fewer contact hours In teachingat times larger classes compensated for this reduction instudents taught, at times no adjustments were made. An-other trend has been a greater emphasis upon providingopportunities for disadvantaged students, As greaternumbers of such individuals are enrolled, the average per-student costs tend to rise because of the larger expenses in-voIved.4 An equally important factor contributing to costincreases has been the growing emphasis upon advanceddegree programs.b a On the average, the preparation of anadvanced degree holder necessitates greater commitment ofresources per student than do expenses associated with thebachelor-level degree. Another and continuing trend hasbeen the acquisition of increasingly complex and ex-pensive equipment, particularly in research-oriented fields.Such needs are a direct function of the knowledge ex-plosion which has necessitated more sophisticated means,of data analysis, storage, and retrieval. This trend will leveloff only in the undesirable event that institutions of highereducation, by failing to maintain modern facilities, begin torelinquish their position in the forefront of the knowledgeexplosion.

In order to complete the picture regarding financiallimitations, factors associated with ever-rising costs mustbe considered in conjunction with the stowing or stabiliz-ing levels of financial support which many institutions arecurrently experiencing. Both commercial and privatecontributions continue to be affected by recent recession as

112

the following excerpts indicate:Total corporate giving to education in 1970declined 9 38 percent to $340 million, About 80percent, or $270 million went to higher education,'

In its most recent report on private gifts to U.S.colleges and universities, the Council for FinancialAid to Education found that last year (1970) thefirst average annual decrease in more than adecade was registered,"

While a modest increase (6.5 percent) in total givingto higher education occurred in 1971, the message is clear.Philanthropic contributions to higher education werebecoming more difficult to obtain."

At the same time, governmental support at both thefederal and state levels has become more tenuous as thevoting public, in part because of recent campus disorders,lodemands a closer accounting for its tax dollars. In addition,one body of thought contends that state governments havereached or are approaching the limits of their capacities tosupport higher education, particularly in view of pressingand ever more competitive demands for slices of the taxdollar.''

The extent and severity of the financial crisis maybest be demonstrated by an illustrative example. TheNatio-J Association of State Universities and Land GrantColleges (NASULGC) recently stated that 14 of its membershad reported operating fund deficits for the 1969.70 aca-demic year.la They further reported that 44 members hadreported increases in their operating budgets of 10 percentor less, while a 10 percent increase was determined to bethe minimum average requirement for matching the effectsof inflation and enrollment expansion."

Nor is this situation restricted to the universitylevels. A recent study has disclosed that many smaller,private colleges were showing operating deficits by 1969(1968-69) equal to 26 percent of the book value of theirassets.'4 In a follow-up study, the author predicted thefailure of as many as 100 of the institutions in the nearfuture, given "business as usuai,"is

Another survey of 48 private four-year liberal artscolleges revealed that most were characterized by worsen-ing financial conditions with no help in sight.15 The re-searchers felt that colleges have experienced more inflationthan the economy as a whole while producing only slightlymore growth.

In many instances where colleges and universitieshave continued to operate in the black, methods for sus-ta:Ning financial solvency have included tuition increasesand boosted rates of admission rather than actual manage-ment improvement), There are strong indications that suchapproaches, wherein the student is asked to pay for dis-parities between productivity and cost, are nearing ex-

haustion." Another factor which must now be faced resultsfrom the declining birth rate between 1951 and 1968," Pro-*Bons indicate that the college-ago population (18.24 yearolds) will actually decline between 1980 and 1985.20Furthermore, enrollments may be expected to slow, wherethey have not already done so, at a much earlier date," Inthe fall of 1972, for example, the percentage of high schoolseniors entering higher education, which had been risingfor decades, actually declined. The prospects of actualdeclines in enrollments present a problem with whicheducational institutions have rarely had to deal. With theexception of enrollment declines which resulted from thegraduation of veterans (about 1950.52) and declinesattributed to enlistment in World Wars, onroilmentpressures have mounted continuously throughout thecentury. If projected decreases to tecur, and if capitalinvestment expansions continue at present rates, onoshould not be surprised to see many Institutions developingover-capacities at least in certain types of facilities by the1980's. Studies show the small institutions are in manyinstances already feeling the pinch for students. A surveyof 431 private colleges for the period 1965.1972 indicatedthat even during this period of generally increasingenrollments, these smaller schools showet: declines inpercentage increase."

In terms of cost it might be expected that fixed costswill be maintained at current levels while variable costswill be proportional to facility utilization. It will be the rareinstitutions which can reduce and control fixed costs atrates commensurate with in pending enrollment declines.

Break-even AnalysisA simplified break-even analysis such as that pre-

sented in Figure 2 represents the fixed costs (FC), variablecosts (VC), and total revenue (TR) lines. if one assumes thediffiouities involved in reducing fixed costs rapidly in theface of enrollment declines and that given any level ofinflation variable costs will not be reducible withoutdamage to educational quality, the question for each insti-tution is whether its enrollment will decline below point E.Point E designates the "break-even" point between profitsand losses.

The third question which should be consideredrelates to the social responsibilities inherent in the role ofhigher education. While social responsibility assumet:multi - faceted dimensions, this question asks to what degreean institution should monitor and align itself with demandsfor the specific degrees which it supplies. An institutionwhich does minitor such demands for degrees should beable to ascertaii what portion of the total market it cansupply. Logically each discipline would endeavor to tailorthe number of degrees granted to anticipated demands.Problems arise when such procedures are conducted cn aninformal and non-systematic basis as is the case ci some

Monlgomory, Janney, McLaughlin and Slog

institutions, and from the competitive desire of depart.ments to attract majors. The impending degree surplus Inspecific areas dictates that suet procedures need to beformalized, In instances where a given discipline findsitself glutted with degree candidates designated for surplusareas, even after accounting for the effects of normal attri-tion and curricular switching, counseling and re-orientationprograms become desirable and necessary, Such a concernalso has economic implications, for example in facultystaffing and in confidence of students and alumni in thepurposes of an institution,

The discussion in this paper thus far has consisted ofa review of current and anticipated challenges facing insti-tutions of higher education now and in the not so distantfuture. The second portion of the paper suggests methodswhich have been either proposed or implemented in re-sponse to the situations outlined. Again it may prove best toseparate considerations into three primary areas: costcutting procedures; income producing methods; and mana-gerial actions. (Admittedly there is considerable overlapwithin the three.)

Perhaps the easiest area to subcategorize is cost-cutting, This might then be loosely divided into the follow-ing categories: selective cuts, across the board cuts, consoli-dation, readaption, and the "swim or sink" approach, TheNational Association of State Universities and Land GrantColleges has compiled a list of seven econymy measureswhich are employed with varying degrees of frequency bymember institutions." Of these seven measures, defermentof maintenance, elimination of new programs, and faculty-staff cutbacks or freezes were found to be most frequentlyemployed, Some of these involve stop -gap methods whichcan only erode the overall educational quality of the schoolin the long run,"

The Academy for Educational Development, Inc. hascompiled a list of some three hundred and nineteen speci-fic actions and approaches aimed at alleviating financialdifficulties. 25 The list deals not only with cost - cuttingmethods but additionally cites several infrequently ex-plored alternatives in the area of income production someof which will be mentioned below.

One method of cost-reduction receiving much atten-tion is to combine small, related departments. Such an ap-proach, it may be argued, eliminates needless dual costsand administrative replication. A more stringent approachalso receiving attention is the painful but realistic elimina-tion of departments which have ceased to grow and haveactually shrunk in recent years.

Consortia represent another concept receiving increased scrutiny." Cost reductions can and do occur whencontiguous colleges form alliances in offering courses,facilities, and services. Successful cooperative arrange-ments currently functioning in Washington, D.C., NewHampshire, Boston, and other areas attest to the benefits of

113

FINANCLAI

DOLLARS

($)

LOSS

Figure 2

BREAK-EVEN DIAGRAM

(TR)

PROFIT(VC)

(FC)

E

consortia.A related method which has long occupied a contro-

versial position among faculty and administration alikemanipulates so-called productivity ratios, such as thatbetween student credit hours produced and full timeequivalent faculty.27 Large lecture sections almost uni-versally are coming into more frequent use. "Work-study"or "special" study course offerings which allow for inde-pendent student research are also felt to hold answers, Theprimary villain is felt to be indiscriminate course prolifera-tion.

While tuition increases and the assessment ofvarious fees remain valid and realistic practices, revenuefrom them should not be viewed as an area of major thrustin the quest for balanced budgets, There remain realisticupward limits to higher tuition, especially in an era when

114

ENROLLMENT

the open door to education promises to become a reality.The expectation of higher income through enrollment in-creases is warranted only in instances where admissionsand recruitment programs have been revitalized and co-ordinated in recognition of the keen and growing compe-tition. Since the problem of boosting income sources maybe expected to become even harder to solve, alternativesmust be sought. Certainly large additions from federal andstate governments should not be expected."

One method of fund acquisition which has receivedmuch attention calls for investments which involve higherthan traditional risks but which also promise greaterreturns.14 One contributor to jellema's Efficient CollegeManagement listed among a college's diverse portfolio ofinvestments federal housing projects, private subdivisions,motels, laundromats, and even a harness racing track. The

same author rotors to the "small town banker approach"which seems to characterize tho views of many directors offinance and boards of trustees.30 Another plan, the so-calledYa lo Tuition Postponement Plan, permits students to delaypayment of portions of yearly tuition in return for a per.cenlago of future income over a specified period, A recentitem in ERIC echoes the familiar but true refrain thatalumni remain a largely untapped source of funds. Thoauthor of it further asserts that trustees might attempt toshoulder more of the responsibility in the area of fund ac-quisition.31

Efficient ManagementEfficient management, the third area for roviow,

offers tho promise of both the greatest frustration as well asthe best hope for survival. Realism dictates that educationalinstitutions anticipate an extended period of financialstringency. While funds in many instances may stabilize,ono cannot expect demands upon thorn to do likewise. Ac-cordingly, one must strive to do moro, or at least' as much,with less. A most fundamental change is needed in thoconcept of educational management, Whether one likes itor not, it appears that continued financial solvency willnecessitate adoption of approaches analogous to methodsemployed by modern commercial management. As much asone may eschew concepts such as product deletion, forexample, this perspective is precisely what lies behindcurrent emphasis upon priorities, Modern managementtechniques and methods should leave no area untouchedby introspection and analysis. Three primary areas withwhich management should concern itself have been out-lined by one author as follows: (1) careful analysis of thorelationship between the utilization of resources and theaccomplishment of goals; (2) quest for maximumeconomies with minimum sacrifices in quality of educa-tion; (3) encouragement of rapid and flexible adaptation tochanges in demands,3a

The realistic response to point one above clearlyInvalidates use of blanket reductions. Yet. as the followingquotation indicates, this procedure is popular.

Across-the-board cutbacks (for instance, budgetslowered by 5 percent for ail departments andschools), or a university wide "freeze- on hiring oron raises, are the comming pattern, . Equityamong divisions whose relevance to theuniversity's main mission WI litiS a groat deal is theopposite of selective pruning.33

If across-the-board cuts are not the answers, prioritic,3 needto be established and each program individually evaluatedin terms of these. Having done so one can confidently andjustifiably reduce costs in less or non-essential areas. (Itwill not be easy, however.)

The unifying thread to the suggestions which havebeen offered is the need for management by objective. In a

Montgomery, Janney, McLaughlin and Slog

time of stringency such as currently exists, It would bybeneficial for each university community to reassess itslongrango goa)i and associated short-rango objectives andthen to mako decisions based upon those objectives. Thoessential element to each daily decision, such as the costreduction discussed herein, Is tho evaluation of its effectupon the achievement of long-rango goats such that short-range actions do not jeopardize the more pervasive goals.

Tho process of evaluating and, if need bo, curtailingthe operations of an academic department in higher education Is analogous to the product deletion decision often en-countered by commercial enterprises, This approach Is re-flected in two product decision proposals prepared on thocampuses of Connecticut and Minnesota, The 'essence ofeach of those proposals is a series of questions with cate-gorical answers which relate a product's compatibility withand position among existing product linos. The questionsand their responses combine to form weighted groupings(for example, marketability and channels of distribution)which are evaluated to determine an index value. Levels ofdegree production and enrollment would enter into theprocess as crucial variables. Such management techniquesdo not eliminate subjective decision making altogether, butthey do provide a narrowed spectrum over which decisionsaro to bo made while providing uniform sets of criteria.

The ObjectiveTho ultimate objective of this presentation is not to

suggest specific techniques which might bo employed tooptimize utilization of limited resources but rather to callfor an overall plan of attack, hopefully before a crisis arises,In light of this objective, two useful planning documents,mentioned above as examples of product decision pro.posals will bo briefly discussed. Tho first document is onowhich was presented by the Long Range Financial PlanningCommittee of tho University of Connecticut to its presi-dent:14 The strength of this proposal Is that it is orientedtoward specific recommendations for improvement in areaswhich tho committee saw as economically deficient. Forexample, in consideration of income and under the title ofinternal pricing, tho committee recommended to theStudent Welfare and Scholastic Standards Committee thata fee for dropping courses after a certain point at the beginning of each term bo instituted. Similarly, the committeerecommended to the Faculty Standards Committee a seriesof ratios for use in evaluating fund allocations to depart-ments.

The specific recommendations suggested in the Con-necticut document, however, represent only part of theimportance of this effort, The importance Iles in the factthat specific recommendations were made for specificbodies or individuals to act upon by a concerned group offaculty and administration.

Another excellent budgeting planning document has

115

FINANCING

been developed at the University of Minnesota.3s 'rho thrustof the Minnosota plan was to design a budget to mako 15percent of the 1971.72 instructional funds avallablo for re-allocation within the university over tho period of 197275.As a part of this planning document, the joint committer.)effort at Minnosota posed many penetrating questionsunder broad headings such as Uniqueness of Programs,Vitality of Faculty and Student Body, and Costs of Program.Theso questions provided areas of concern from the generalto the specific within which decisions were to be made andactions undertaken. 'rho areas ranged from the breadth of aconsideration of program effectiveness and redundancy tothe depth of a concern about hidden costs rolativo to anyparticular program or department. Such questions, tho coroof tho Minnesota plan, provido the Framework for planningand budgeting at all levels in the institution. 'rho samequestions could be employed by any institution of highereducation but would bo answered differently based upontho goats and objectives of each. The final segment of theMinnesota plan presented actual goals in terms of per-centage reallocation of funds for each year over the three-year planning period. For the first year of this three-yearperiod, each budget unit was to reduce its budget by 6percent. The unit or the college could then mako proposalsto retain half this amount (3% ) for new or renewedprograms. The other 3 percent was reserved for use by the

total University to support now programs. Horein lies thostrength of tho plan, for the univorsity began with aproblom (financial difficulties), assessed its organizationalgoals, developed a broad freamowork for continualplanning, and finally established objectives by whichindividual actions should bo guided and ()valuated,

When one compares tho two documonts proviouslydiscussed, the similatities in planning processes becomoclear. Both planning documents resulted from the mallzation of a problem and the development of specific actionsto remedy it based upon organizational goals. Also, each oftho plans presented a unified approach to the organi.zational problems rather than a piocemeal solution toresults of an underlying problem,

In summary, conditions aro either present or uponthe horizon which may retard financial growth, As indi-cators emergo which suggest a need to hold or to reducobudgets, tho most useful approach appears to bu to mako .athorough self-analysis as a basic for reorientation andimaginativo plans for tho future. Limited measuros such asdeferring maintenanco or purchases usually provo ineffec.tivo over a period of dm. Moreover, cooperation will boneeded among the administration, faculty, and students inorder to find the bost solutions whilo avoiding to the extentpossible loss of moralo and effectiveness. Clearly, therewill be a need to plan and to manago effectively.

National Center for Educational Statistics, Projections of Educational Statistics to 1979-80 (Washington. D.C.: U.S.O.E., 1971), page 3.

Campbell Geeslin, ed., 'Total Cost of Higher Education Continues to Rise," College Management, Vol 6, January, 1971, pp. 10-12.

A Fact Book on Higher Education (Washington, D.C.: American Council on Education, 1st Issue, 1971). p, 71.20.

4 Geeslin, "'Total Cost of Higher Education Continues to Rise," p. 11,

' Proceedings: A Symposium on Financing Higher Education (Manta: S.R.E.B., Juno 12, 1989). p. 20.

01 Geeslin, "Total Cost of Higher Educat on Continues to Rise." p. 12.

' "Occupational Education Bulletin" (Washington, D.C.: American Association of Junior Colleges. Vol, 7, January, 1972), p. 2.

Graven Hudgins and Ione Phillips, People's Colleges in Trouble (Washington, D.C.: NASULGC, July 0, 1971). p, 18.

I Campbell Geeslin, "Less Money is Available to Higher Education," College Management, Vol. 0, January, 1971, p. 13: "Report SecsNew Trends in Giving," Chronicle of higher Education, May 27, 1972.

Symposium on Financing Higher Education, pp. 29, 31.

Hudgins and Phi/lips. People's Colleges In Trouble, p. 1.

Ibid., pp. 21.22.

11 A Fact Book on Higher Education, 2nd Issue, p. 71.77.

14 William jelleina, "The Red and The Black. Special Preliminary Report on the Financial Status, Present and Projected, of PrivateInstitutions of Higher Learning" (Washington. D.C.: Association of American Colleges, 1970) ERIC ED 058837.

15 William 'enema, "Redder and Much Redder: A Follow up Study to The Red and the Black (Washington, D.C.: Association ofAmerican Colleges, 1971). ERIC ED 058 838.

14 Hans H. Jenny and G. Richard Wynn. The Turning Point. A Study of income and Expenditure Growth Distribution of 48 PrivateFour-Year Liberal Arts Colleges, 1960.1970 (Wooster, Ohio: College of Wooster, 1972) ERIC ED 065 019.

116

Montgomery, Janney, McLaughlin and Slog

t Commission for Independent Colleges and Universities. Update Study of the Financial Condition of independent Higher Education in thuCommonwealth of Pennsylvania (Washington, D.C.: McKinsey and Co., February 1972) ERIC El) 060 812,

la William !enema. "The Numbers Game, A Study of Enrollment Patterns In Private Colleges and Universities" (Washington. D.C.:Association of American Colleges. January 1972), ERIC 058 846.

14 A Fact Book on Higher Education. 2nd issue, p. 71.69.

Proiec lions of Educational Statistics, p. 22.

II Hudgins and Phillips. People's Colleges. pp. 4-5.

foltema, "Redder and Much Redder,"

," Editorial Protects for Education, "15.Minute Report," (Washington, D.C.: Vol. 7, May 1, 1971) pp. 1.4.

44 "glop-Gap Measures or Fundamental Change?" Finoncing Higher Education (Atlanta: S.R.EB., No. 24, 1971) p. '3.

"Ways Colleges and Universities Are Meeting the Financial Pinch," Management Division, Academy for Educational Development(New York, New York: April 1971).

David A, Trivet, "Small College Management: Key to Survival" Research Currents, ERIC, George Washington University (March1973).

0 ibid.

sa " For Your Inform otion," (Washington, D,C.: NASULGC, Office of &scorch and information, Circular No. 179, March 2, 1973)

29 Trivett. "Small College Management;" Sharvy Umbeck, "New Approaches to Finonce" in William W. /Memo, ed., Efficient CollegeManagement (San Francisco: jossey-Bass. 1972), pp. 122-28.

'4' Ibid.

Trivelt, "Small College Management,"

I, Ibid.

"Mark E. Stern, Marketing Planning: A Systems Approach (New York: McGraw-Hill, 1966) pp. 52-57.

" The Long Range Financial Planning Committee of the University of Connecticut, "Approaches to Economy in Operation at theUniversity of Connecticut: Report of the Committee," February, 1972, pp. 1.14.

41 University Senate Committee on Educational Policy and University Senate Committee on Resources and Planning of the University ofMinnesota, "Accountability and Educational Criteria: University Planning for Selective Growth," July 22, 1971, pp. 1-22.

117

UNIVERSITY IMPACT ON LOCAL ECONOMY

Bernard S. Sheehan and Barbara Serediak, University of Calgary

Universities influence their local economies bygenerating additional gross real goods and services frominstitutional expenditures and personal expenditures ofstaff, students and visitors. A simple minimum costmethodology for delimiting the economic impact of auniversity on its local economy is described and illustratedby The University of Calgary's impact on the City ofCalgary for three fiscal years.

MethodologyThe methodology includes a calculation of total cash

flow into the local economy from four types of university-related expenditures leading to estimates of economicimpact defined by local jobs and by earnings directly orindirectly supported by the university. The proposedmethodology Is structured to take full advantage of thepeculiarities of the local economy in order to minimize thecost and time required to use it. Therefore, some judiciousadaptation may be necessary when the methodology isapplied to different types of local economies.

Staff disposable income, net university expen-ditures, visitors' expenditures and student expenditures areinitial local income as a hown in Figure 1.

Local staff expenditures are assumed to equal staffdisposable income, Such items as savings and privateexpenditures outside of the city are assumed to be offset byother income sources. Surveys which would estimateleakages are costly to produce and the net ettect of thisapproximation is Judged not to be significant for thepurposes of this study.

Net non-salary university expenditures include non-salary operating, non-salary trust fund, and capital expen-ditures. The value of goods and services purchased by theuniversity which flow in first-round expenditures into thelocal economy is estimate& based upon the assumptionthat payments to firms flow directly into the local economyin the first-round if the firm has a local mailing address or alocal business telephone number.2

Expenses of non local delegates to campus confer-ences are considered to represent the majority of visitors'expenditures. An estimate of per visitor total local expendi-tures is applied to the number of annual non-localdelegates to campus conferences.

118

Local expenditures per full-time student are basedupon estimates from files of university financial aidoffices.3 It is assumed -that full-time students if they are notattending university would be idle resources or would betaking employment opportunities from someone elsemaking the net effect nil. Also It is assumed that in theabsence of the university, students who live at home wouldhave gone elsewhere to study and further that room andboard expenditures of students living at home are the sameas those for students living away from home .4 The studentexpenditure estimate (not including tuition) is multipliedby the number of full-time students.6

Expenditure and Income MultipliersUniversity related expenditures within the region

create local income, jobs and business investmentopportunities. The first round income accruing to local resi-dents is partially rospent in businesses round after round,becoming an infinite geometric progression creating addi-tional income and employment.

Expenditure income and initial income multipliersare estimated based upon existing studies of various typosof expenditures and regions with similar characteristics.6Multipliers vary depending upon comparative reliance onimports, spending and saving preferences of residents,patterns of consumer spending, and regional industrial andbusiness structures. Multipliers within the ranges of valuesdetermined by relevant studies are chosen as reasonableestimates to multiply by initial income and universityrelated non-salary expenditures within the region.

Employment OpportunitiesUniversities support many jobs in addition to those

on campus. These jobs are an indicator of the economicimpact of the institution on the community. An estimate ofjobs supported is obtained using the average of results oftwo methods.

Method I University employment Is assumed toform part of the regional economic base which consists ofthose activities providing employment and income onwhich the local economy depends. This theory' premisesthat "nonbasic" jobs are generated through employment ofper§ons in activities which form part of the economic base.

Sheehan and Sarah kik

Figure 1

INITIAL UNIVERSITY-RELATED EXPENDITURES

STAFF NET

DISPOSABLE NON-SALARY VISITORS' STUDENTS'

INCOME UNIVERSITY EXPENDITURES EXPENDITURESEXPENDITURES

100%DISPOSABLE

INCOME

X%LOCAL

INCOME/X%

LOCALINCOMEi

(100-X) % COST OFMATERIALS, NON-LOCAL INCOME\

(100.X)% COST OFMATERIALS, NON-LOCAL INCOME

119

UNIVERSITY IMPACT

Several estimates of the size of the multiple effect ofadditional basic employment have been made for varioustypes of expenditures and regions. Employment multiplierswithin a region are laigui the greater its population andemployment diversity, and the less its dependence onimports.. A value for the employment multiplier may bebased upon studies of regions of similar size, diversity andinterdependence, and upon studies of similar types ofexpenditures, This value multiplied by full-time equivalentuniversity employees yields an estimate of additional sobssupported by university wages and salaries.

In addition, the nonsalary operating, non.salarytrust fund, and capital expenditures of the universitysupport employment. Ratios of the value of constructionwork and wholesale trade to the number of employeeswithin each industry are applied respectively to universitycapital and nonsalary operating and trust fund exprndilutes to yield initial employment estimates. Averageemployment multipliers for contract construction andwholesale trade are estimated based upon studies of similarregions, and are applied respectively to these initialemployment figures.

Method II The number of local sobs attributableto the presence of the university may also be estimated byapplying an "employment requirements coefficient."Coefficients measure changes in local employment

120

associated with the average household dollar spent locallywhen direct and indirect production requirements andinduced income effects on local production are taken intoaccGunt.v A suitable range of the coefficient is suggested byCaffrey and Isaacs in terms of man-years of employmentper eollat of university related expenditures in localbusinesses. A coefficient which is chosen based uponcriteria similar to those of the income and employmentmultipliers is multiplied by staff expenditures and universky non.salary operating, non-salary trust fund and capitalexpenditures,

Mathematical Statement of the MethodologyThe methodology is summarized algebraically in

Figure 2. Equation I shows the calculation for initialbusiness volume while income generated is given byequation 2. The two methods of estimating employmentgenerated are given by equation 3.

Estimating Trend of Future impactAssuming that university related expenditures are

proportional to full-time students, and that initial incomemultipliers in the projection year equal those in the baseyear, annual local income generated over succeeding yearsmay be estimated based on student projections. Thisestimate is likely to be low if the local economy is expand.

Figure 2

ALGEBRAIC STATEMENT OF THE METHODOLOGY

Equation 1

B W + h (T + C) + fV + gS R

in which:B =I Initial local business volume

W = disposable income of faculty andstaff less that of faculty on leave

h = Locatdisbursrmentstotal disbursements

T = non-salary operating and non-salarytrust fund expenditures

C = capital expenditures= per visitor estimated local

expenditures

V cm number of conference delegates

g = per full-time student estimatedlocal nontuition expenditure

S = full-time studentsR = revenues from university

facilities

Equation 2

I aW +11 (B W)in which

I = local income generateda = estimated Initial income multipliers

= estimated expenditureIncome multiplierEquation 3

JL = A (E + mhC + nhT) ; = E + p (W + h (T + C))in which:

FTE jobs using Method I

= FTE jobs using Method it

A = estimated employment multiplierE = FTE faculty and staff employed by university

m = F construction employeesvalue of construction work performed

n = FTE whotesate trade employees

p = employment requirements coefficient

ing because values of multipliers in the projection yearlend to be higher than those in the base year.

The proportion of local jobs supported in a givenyear may also be es"ated by assuming that the ratio ofprojected full-time students has a direct relationship to theprojected regional employment directly or indirectlysupported by the university,10 An estimate of full-timeequivalent employees within the region in the given timeperiod is used to determine the estimated proportion oflocal lobs supported by the university.

Discussion of MethodologyThe economic indicators of local income generated

and full-time equivalent jobs created are not the only oneswhich could be chosen.31 A balanced picture of the totaleconomic impact of the university on the community mustinclude many dimensions such as expansion of local creditbases, unrealized local business volume duo to the univer-sity having its own facilities, the proportions and value ofcapital and properly which relate to university-generatedbusiness volume, and local government expenditures andrevenues duo to existence of the university. Each of thosefactors are important yet require detailed analysis. Theiromission limits the results and the significance of theiromission must be judged in light of the purpose of each

Sheehan and &reit lak

application of the methodology. The proposed methodologyattempts to yield evidence of the economic impact of theinstitution on the local economy to assist the institution inproviding an accounting for tes,.ucces piuvided by govern-monis, Thus, local income generated and full-timeequivalent jobs created seem most relevant for thispurpose,

Figure 3 shows the Income and jobs generated in theCity of Calgary by the University of Calgary in 1988/87,1970171 and 1971172, Thus, as the number of full-timestudents increased from 4074 to 9173 the local income and-butable to the university grew from $21 to 71 million andthe number of full-time equivalent jobs resulting fromuniversity-related activities jumped from 2850 to 7990.

Needs of people whose income and jobs aresupported by the University of Calgary provide substantialmarketing opportunities for local retail and serviceestablishmerns. By assuming that universitygeneratedincome will be spent in a manner similar to that of averageCanadian familles,u distribution of local retail activity infood, housing, clothing, automobile, furnishings and otherindustries can be estimated, Figure 4 shows the distributionof income generaged by 1971172 University of Calgaryrelated expenditures into these sectors.

Figure 3

RESULTS

IMPACT OF THE UNIVERSITY OF CALGARY ON THE ECONOMY OF THE CITY OF CALGARY

FULL-TIMEYEAR STUDENTS

GOVERNMENTINVESTMENT(S'000,000)

INITIALLOCALBUSINESSVOLUME(S'000,000)(11),

LOCALINCOMEGENERATED(V000,000)(I)

FIE JOBS(ii + J1)/2

ESTIMATEDPERCENTOF CALGARYEMPLOYMENTSUPPORTED

1966/67 4074 23 24 21 2900 31970/71 9237 42 59 64 7100 51971/72 9173 51 64 71 8000

121

UNIVERSITY IMPACT

Figure 4

DISTRIBUTION OF LOCAL 1971/72 INCOMEGENERATED BY

UNIVERSITY OF CALGARY RELATED EXPENDITURE

TOTAL INCOMEn: $71 MILLION

FOOD& BEVERAGES

12.2

(MILLIONS OF DOLLARS)

*The authors wish to express their appreciation to University of Calgary, for her contribution in the develop-Mrs. Margaret Reti, Office of Institutional Research, The ment of the methodology described in this paper.

122

Sheehan and Sorehlah

This estimate does not include investments, bank transfers, travel expenditures, pensions, fellowships, bursaries. scholarships orstudent loans.

In order to minimize costs of dMi preparation. the ratio of local to total university expenditures calculated for one time period may beassumed constant and multiplied by capital. non salary operating and nonialary trust fund expenses of succeeding limo periods to yieldestimated local firSi.round university expenditures,

In the University of Calgary illustration, a questionnaire calculated by the Student Affairs Office to a ten percent sample of full.timestudents verified the estimates of per student expenditure used,

4 This as assumption was tested with University of Calgary data to estimate the decrease in total students' expenditure if room andboargl expenses of students who live at home were not Included. Room and board payments less an estimate of the costs to families for Ihosostudents were deducted. The difference in cash flow into the local economy was approximately three percent, Therefore, the extent ofeconomic impact was not significantly changed,

$ University housing. food services centres and bookstore revenues are subtracted from the total cash flow because gross expenditures inuniversity facilities are included in university operating expenses.

Several estimates of expenditurincome multipliers which may be applied to university expenditures are cited by John Caffrey andHerbert H. Isaacs, Estimating the Impact of a Co liege or University on the Local Economy (Washington, D,C,1 American Council onEducation, 1971), p. 44.

Charles M. Tiebout, "Tho Community Ecunurnic Base Study," Supplementary Paper No, 16 (New York: Committee for EconomicDevelopment, DeCember 1962).

John Caffrey and Herbert H, Isaacs. p. 44. City of Calgary employment Is approximately 1/5 in the manufacturing sector and 2/5 in thetrade and service sectors, .Calculated from: "Estimates of Employees by Industry," Alberla Business Trends (Edmonton: Alberta Bureau ofStatistics, fanuaryDecember 1971 issues).

w fohn Jaffrey and Herbert H. Isaacs. p. 44.

10 A sensitivity lest was performed on this assumption using historical University of Calgary and City of Calgary data, The test yieldedratios only slightly higher than those calculated using this assumption.

II Another university impact is changes in nearly relative real estate prices. A comparative analysis of residential lots and houses atvarious distances from the university was done. Ratios of selling prices to assessment values in dose and distant areas were compared forhouses with similar living areas and lot sizes. The analysis shows that the university may have inflated selling prices of proximate homes. Thelocation of The University of Calgary in a new section of a rapidly expanding urban region may make this observation peculiar to universitiesin this sort of physical environment,

Dominion Bureau of Statistics, Urban Family Expenditure, 1967, Catalogue #62.530 (Ottawa: Queen's Printer, March 1971) p. 311eilmonton pattern).

123

THE USES OF COST DATA IN INSTRUCTIONAL PROGRAM EVALUATION

Higher education finds itself in a rapidly changingworld in virtually all dimensions from the basic economicsof financing to patterns of demand for programs of study, Aleveling off and projections of actual decline in enrollmentsfrom the "traditional" 18.24 ago student group arc beingheard with increasing frequency, A multitude of now innowitty() educational programs mining from the use ofcomplicated multimedia technology to the "UniversityWithout Walls" and degree by examination are being pro-posed and tested. Institutions find themselves facing un-precedented financial hardships and many proposals forchanging basic methods of support for post-secondaryeducation, Faculty and staff union organizational effortsare increasing. Student tuition and other costs aro risingand student interest in administrative affairs is increasingat at least a comparable rate.

The consequences of such dimensions of change arewell known to those concerned with management at institutions of higher education: increasing pressures from stateand federal agencies and from alumni for "accountability:"increasing pressures from students for "Relevance" (how-ever this may be defined); and increasing pressures fromall funding sources for "Efficiency." The sum of all this isthat the instructional process of programs, both traditionaland new, are coming under pressure for review as neverbefore,

Instructional Program Reivew, an Economist's ApproachI propose that the basic criteria around which the re

view of any Instructional program be built be that there is ademonstrable demand for the outcomes of the program.That is, the establishmnt of a new instructional program orthe continuation of an existing fine should rest upon evi-dence of a significant and reasonably consistent demandespecially with respect to the next decade,

There are two dimensions to this really very basicproposition: the first concerns "student demand" and thesecond concerns the "social demand" or the needs ofsociety in general.

The most Important dimension must be studentdemand and for students it can be assumed the demand isa function of:

124

Robert D. Lamson, Boeing Computer Services

1. The extent to which any educational program results in"certification" which is not achievable in anotherprogram and which is necessary to achieve some post.educational goal (usually, but now always, employ.ment); or

2. the degree to which the program is unique and offersan educational experience which is stricly for its consumption value; or

3. the degree to which the program provides importanteducational support or "service" to other programswhich are in great student demand.

In assessing student demand it must be rememberedthat in all public Institutions of higher education, and mostprivate ones as well, the student is required to pay considerably less than the actual variable costs associated withhis educational program. This is due to the fact that institutional subsidies, whether derived directly from taxes orultimately from "tax supported" endowment funds, resultin substantially lower and generally more evenly distrlbuted student borne educational costs than would be themasa in an environment where all or most of the variable in-structional costs would be charged to the student, An advantage of the latter situation, which would certainly involve some significant student aid, Is that the mere exis-tence of any program in itself would imply "sufficient"student demand. However, In the situation where costs tothe student are significantly below instructional costs andwhere the student-borne costs do not vary appreciably ac-cording to actual differences in instructional costs acrosseducational programs, a different pattern of studentdemand emerges. Those programs which are relatively highin actual instructional cost enjoy relatively higher studentdemand and those with relatively low instructional costhave rotatively lower student demand than the case inwhich student Fees would vary in relation with differencesin longrun variable instructional costs.

In the former situation, without some reviewprocess. it would not be unexpected to find a larger menuof educational programs (some perhaps very small andperhaps high in cost), since there is not differential burdento the student for experimentation on the part of the institu-tion in offering small and/or specialized programs and con-

tinuing even those which prove "unsuccessful" In terms ofstudent interest. In contrast, In the case whom student costsare directly related to actual instructional costs one wouldexpect to find no student demand at all for those programsfor which size or some other aspect carries unfavorable costimplications, and where the educational result of theprogram, whether certification of consumption in nature,could be obtained almost as well in other lower costprograms.

Evaluation of student demand should explicitly takethis fact into account since, ceterIs paribus, relatively lowinstructional cost programs would be expected to oxpandand relatively high instructional cost programs would beexpected to contract under a more "normal"-(in the econo-mic snese) pricing and financing scheme.

Of course the dimension of public need or socialdemand is important since it is presumably this on whichany subsidy is justified, Any instructional program mayproduce useful outcomes which do not accrue to thestudent, These are generally said to be research or publicservice in nature and such outcomes must be weighed injudging ultimately the value of any instructional program.Evaluation in this regard should attempt to identifywhether or not the social outcomes vary in significance ac-cording to the identified differentials in subsidy, composedof the difference between student borne costs and actualinstructional costs as discussed below.

Another dimension which relates very much tostudent demand, expecially in the sense of "potential"demand might be termed program "effectiveness."

Three critical dimensions of effectiveness are1. The success of students who complete their educa-tional program (vis-a-vis those who do not) in achievingsome post educational goal, usually employment;2. The percentage of students who ultimately completethe program; and3. The average and variance around the average of lengthof time required for completion.

The considerations of effectiveness are important be-cause student demand as viewed in terms of numbers ofstudents in any one time period represents only a snap shotof programs. Programs which are identical on a snap shotbasis may be very different in the long term demand/effec-tiveness sense,

These factors student demand, social demand andprogram effectiveness shoutd be considered in the processof evaluating instructional programs in higher education asit generally exists today, with subsidy schemes whichlower and more evenly distribute student borne instruc-tionai costs so as to alter the "natural" pattern of studentdemand.

The Role of Cost AnalysesHow then does cost analysis relate to the very corn-

Hobert D. Samson

plex process of instructional program evaluation as dis-cussed above? I propose that there aro really Iwo roles, eachof which requires a different kind of cost information,

The first role is to determine the ranking of depart-ments on the basis of what student borne costs would bo ifderived on the basis of actual differences in instructionalprogram costs, What doos this mean? means that to evalu-ate the extent of student demand for any instructional pro-gram, the relative "potential price" position of the programin the above sense should be considered and the intensitywith which evaluation criteria are applied should be indirect portion to the potential price ranking of the depart-tnent, That is, high priced programs should be scrutinizedmore intensely than relatively low priced ones in attempt-ing to assess:1. What factors underlie existing student demand, toanswer the question, "What would happen to studentdemand (increase, decrease remain about the same) ifstudent tuition and fees were charged to reflect relativeprogram unit costs," and2, Defining and measuring the extent and value of thesocial outcomes associated with the instructional process.

In essence this amounts to "subjectively simulating"what would happen to program student demand shouldstudent-borne costs (prices) reflect program costs, andevaluating the price-reducing institutional subsidies on thebasis of the extent to which the instructional program ap-pears to produce significant social outcomes.

The next question then is what costs should be in-cluded in constructing relative potential price rankings ofinstructional programs. The answer I feel is relatively easyin concept but difficult in practice. Conceptually, the costswhich should be included are all long-range variable coststhat would wind up being charged to students if they werethe only source of instructionally related operating revenue,In practice we are talking about long run student average or"unit" variable costs, Wherever possible such costs shouldbe computed on more than one year's experience to gainstability in the estimates and will include:

Direct instructional costs based on course-relatedcosts and an induced course toad matrix to translatethese into a student base for each instructional program;

Allocation of all other variable costs associated theinstructional process, with the allocations on the basismost consi,Aent with the nature of their variability (in sofar as possible on the basis of actual user population);

Allocations of quasifixed costs (Lc, fixed in theshort run) including do estimate for annualized capitalcosts on a similar (user) basis or simply prorateddepending on their nature,

If in fact institutions of higher education were tomove to a full or proportional variable cost pricing policy,this is probably how in general student prices would bedetermined. Resulting potential price estimates must then

125

USES OF COST DATA

be "interpreted" and integrated in the review process with:Evidence of program uniqueness and the availability

of "substitute" programs al the same or other institutionswith a lower potential price;

FAidence of the apparent necessity of completing theprogram to receive the educational benefits;

Estimates of the length of time required forcompletion of the program; and

Definitive evidence which is available concerningthe social outcomes of the program,

The results of a review which integrates informationon demand, effectiveness and cost (in the potential pricesense) will generally be tentative decisions to expand,reduce, create or even in some instances terminate educa-tional programs. Once this state is reached a different kindof cost information is required to ultimately reach and seethe consequences of the final decisions. This is programrelated cost information, which identifies the cost implications of changes in the activities of the various academicand other units within the institution caused by the changein status of the educational program under review. Suchprogramrelated costs would include:

Incremental course cost changes (if any) in theparent department associated with changes in the educaHone' program;

Incremental costs in the parent department's Re.search and Public Service activities (if any) caused by thechange in status of the educational program underreview;

Incremental course costs changes (if any) in otherdepartments cuased by a significant change in the "ser-vice load" impact of the new status of the program underreview;

Incremental Academic Supports costs (if any) cause.'by such things as changes in the activity level of librarhthe computer center or other such areas providingacademic support to the instructional, research andpublic service activities of the institution: and

Incremental Student Service (such as housing, foodservice, counselling, etc.) and institutional Support (e.g.executive management, general administrative services,logistical services, physical plant operations, etc.) costsassociated with changes in the status of the educationalprogram under consideration. (Unless the program is amajor one involving a large number of students, theincremental impact upon institutional support is likely tobe minimal).

The point to be emphasized is that the kind of infor.mation required to assess the resource impact of changingeducational programs, either upward or downward, are thecosts associated with the incremental changes in the vari-ous activities involved in or related to the educational pro.gram under review, This is not the studentbased potentialprice (long-run unit cost) information described previously

126

and useful for the early stages of review In assessing thesignificance and stability of student demand and therelation of social outcomes of the program to the Impliedprice subsidy involved, It Is very important that those twohinds of cost Information not be confused and used inter-changeably and thus incorrectly,

The relevant incremental programrelated cost infor.mation is generally very difficult to derive since it requiresamong other things identification of:

Where students in the program Under review comefrom or go to with a major change in program status, andwhat are the course, academic support, student serviceand institutional support changes in level of activityimplied by such a shift; and

What is the fixed/variable nature of the cost associaled with those course or other activities so impacted (heroconsideration may be required of such things as thttenure status of faculty and union agreements).

There is n) simple formula and a great deal of"judgement" is required in making such calculations, Also,it must be emphasized that "allocation" in the traditionalsense, as associated with student based "unit costing" isnot relevant, and only those program costs which can beidentified as changing due to a change in the status of theprogram under review should be included,

Summery and RecommendationsIn summary the following points have been made:

1, For the evaluation of educational programs in highereducation today, there are two kinds of cost informationwhich are required,2. In the evaluation of student demand, program effec-tiveness and the social value involved with the instruc-tional process pricing cost data is appropriate, Thisamounts essentially to long-run student-based variable or"unit" cost information, which includes allocations of allrelated costs including support costs, even those qhich arequasi-fixed in the short run.3. Once the evaluation of the program is made and pro-gram changes appear to be indicated, a different kind ofcost data, INCREMENTAL PROGRAM COST information,is required to judge the resource impact of the potentialchanges.

Institutions, state coordinating bodies or other agencies concerned with the evaluation of instructional pro-grams must have the capabilities to estimate both kinds ofcosts, and at a level of disaggregation which will not arti-ficially gloss over significant differences In either potentialprices or program costs. This requires the capability tocalculate direct instructional costs both by course and bystudent as a base. It requires the capability to identify"induced" loads on a student major/course and instruc-tional program/department basis. Finally, H requires theability and effort to identify the "nature of variability" of

Academic Support, Student Sorvice and InstitutionalSupport costs as well as those related to tho Instructionaldcpartments themselves.

'Moro are no standard formulas or models which can

Po ON

Robert D. Lamson

replace judgement in such cost analyses. On the otherhand, the proper data baso and calculation capabilities arorequired for Informed ludgomrot to bo possible in theprocess tif Instructional pmgrnm evaluation.

127

A NEW LOOK AT TENURE

Thomas M, Freeman and Joseph C. Rossmeler, Michigan State University

Tenure, a term that suggests both sinecure and ace-demlc freedom, continues to undergo scrutiny under theauspices of much publicity. In recent months almost everyprofessional periodical on higher education has devoted atleast ono article to the subject of tenure, The New YorkTimes, the Wall Street Journal, and countless lesser knownnewspapers are also carrying articles about the debate.There is also the proliferation of such books as The TenureDebate, Tenure, Aspects of Job Security on the ChangingCampus, and Tenure in American Higher Education thatare devoted entirely to the issue of tenure. On the locallevel many institutions (e.g., University of Utah, Wisconsin,Michigan State University, etc.) have spent innumerablefaculty and staff man-hours on committees studying thenature of tenure. On the national scale the Commission onAcademic Tenure in Higher Education, sponsored by theAAUP and thr Association of American Colleges, has justconcluded a two-year study. Predictably, along with a fewsubtleties graced by conventional laments about the abusesof tenure, most analysts continue to wave the Americanflag with both hands while arguing for the survival oftenure with a few reforms, of course, on the ground that itis a necessary undergirding of academic freedom.

Since the term tenure has already undergone itsshare of transplants, biopsies and, in some cases, evenautopsies, this analysis shall refrain from further attempt-ing to justify the existence of tenure. Instead, we will lookfirst at several tenure trends in American higher educationand, second, at several braoder issues in higher educationwhich have the potential for affecting tenure. We will posenumerous questions as to whether current tenure trendscan be supported in the future. Third, several solutionsresolving the current tenure crisis are offered, including aproposed new basis for analyzing tenure data.

Current Tenure TrendsThe first question posed concerns whether rank

distribution and tenure are currently at a dangerous level.For this analysis a "dangerous level" is defined as a situ-ation where fixed commitments of institutional staff re-sources have created inflexibility, with program changedifficult and perhaps nonexistent. Tho Commission onAcademic' Tenure in Higher Education (1973) recently

128

slated that higher education is, in academic jargon,"tenured In:" others in recent months have called thecurrent crisis the "tenure squeeze." In a very general sense,the Commission's report concludes that about half of thehalf million faculty hold tenure, with tenured faculty rang-ing from 25 percent at some institutions to 80 percent atothers, with most institutions tending toward the lattersituation.

Probably some of the more reliable positionindicators of tenure trends are facultyby rank distribution,potential faculty eligible for tenure, and actual tenuredfaculty rates. Tenure is not always related in direct pro-portion to rank, therefore the percentages of faculty holdingthe rank of associate or on occasion, the rank of full pro-fessor is not necessarily indicative of the actual percentageof faculty holding tenure. Three procedural circumstancesaro the cause of these discrepancies. First, some institutionsaward tenure to instructors and assistant professors uponreappointment after the successful completion of a pro-bationary or term period. Second, some institutions issueappointments to the top two ranks on a probationary orterm basis rather than give automatic tenure. Third, thebase for the computation of the percentage of faculty ontenure varies considerably among institutions.

Rank DistributionAside from these cautionary limitations one can esti-

mate, with a certain degree of accuracy, trends about tenureby looking at the rank distribution, For example, Table Oneindicates the percentage of faculty members at the fourbasic ranks in land-grant colleges and state universities forthe years 1951.52, 1961.82, and 1971-72. From these data onecan conclude that as the rank distribution of faculty movesfrom instructor to professor, more faculty become tenured.Furthermore, one can assume, that since the percent of pro-fessors has increased by 8.5 percent, the number of tenuredfaculty has probably increased at least an equal factor.

The trends which show the rank distributions ofstate universities and land-grant colleges differ somewhatfrom distributions of a more heterogeneous group of institutions, For example, Trow (1973), in his large nationalsurvey of faculty in a random sample of over 300 insti-tutions in 1969, found the following rank distribution:

Freeman and Rossmolor

Table 1

PERCENT OF FACULTY MEMBERS AT FOUR BASIC RANKS IN LAND-GRANTCOLLEGES AND STATE UNIVIERSITIES, 1951.52, 1961.62, 1971.72

Rank 19512 1961.62 b 1971-72'

Professor 26,7 27.8 33.2)47.7 )52.1 )58.2

Assoc. Professor 21.0 24.3 25.0

Ass't Professor 28.7 28.4 30.8)52.4 )47.9 )41.8

Instructor 23.7 19.5 11.0100.1 100.0 100.0

a) 68 Land-Grant Colleges20 State UniversitiesData from Faculty Salaries in Land-Grant Colleges and State Universities,federal Security Agency, Office of Education, Circular No. 358,19!'1.52.

b) 57 Land-Grant Colleges82 State UniversitiesData from Salaries Paid and Salary Practices in Universities, Colleges,1961-62. Higher Education Series, Research Report 1962-R2, Researchdivision, National Education Association, February, 1962.

c) 57 Land-Grant Colleges40 State Universities which Includes numerous state university systemssuch as Wisconsin, CUNY, and California, etc. Faculty data were takenfrom the "Annual Report on the Economic Status of the Profession,1971.72," AAUP Bulletin, 58 (2) Summer. 1972, pp. 201-293.

ProfessorAssoc. ProfessorTotal

Ass't ProfessorInstructorTotal

25.4% all for 1971-72. The Big Ten, comprised of institutions24.0% which are highly graduate-program oriented, have the49.4% highest percentage of upper-ranked faculty.

30.5%20.1%50.1i%

Evidently, the more graduate-oriented the sample ofinstitutions, the more likely the upper ranks will hold ahigher proportion of the tenured faculty. This inference isstrongly supported by figures In Table Two, which exhibitsthe rank distribution of 1,244 institutions listed in theAAUP annual survey, the rank distribution of 130 graduateinstitutions included in the recent ACE study (Rouse andAndersen, 1970), and the rank distribution of the Big Ten,

Faculty Eligible for TenureThe trend of tenured faculty is also directly affected

by the number of faculty eligible for tenure, i.e., in thetenure stream.' A current O(R tenure study at MichiganState University' shows that of 62 universities who re-sponded, faculty eligible for tenure in the four ranks in-creased 5.1 percent between 1970.71 and 1972.73, as shownin Table Three.

Furniss (1972) found that of the 413 institutions hesurveyed, 33,3 percent award tenure to instructors and 85.2percent award tenure to assistant professors. Shaw (1971)

129

TENURE

Table 2

COMPARISONS OF FACULTY RANK DISTRIBUTION FOR 1971.72

Rank 1,244 Instit.of AAUP

130 ACE' Big Ten

1110.

Professor 25.3% 33.4% 36.4%)49.8% )58.6% )60.8%

Assoc. Professor 24.5 25.1 24.4

Ass't Professor 35.6 31.3 27.9)50.2 )41.4 )39.2

Instructor 14,6 10.4 11.3

100.0 100.0 100.0

*Source: "Annual Report on the Economic Status of the Profession, 1971-72." AAUP Bulletin, 58 (2),Summer, 1972, pp. 201-233.

a Lawrence Institute of Paper Technology, Peabody College, and the University of Delaware, arenot included. Figures Include the entire University of Wisconsin and the University of CaliforniaSystems.

determined in his study of 80 stale universities and land.grant colleges that 46.3 percent offered tenure to assistantprofessors and 27.5 percent offered tenure to instructors.Awarding tenure to the lower ranks is not necessarily arecent phenomenon, for Pfinster (1957) learned in his studyof 128 colleges that 34.4 percent award tenure to assistantprofessors and 44.5 percent to instructors,

Faculty Tenure RatesActual faculty tenure rates are still the best indicator

of tenure trends. Previous research also shows that In mostinstitutions about half the faculty hold tenure appoint-ments; however, the range of variation Is quite extensive,Shaw (1971) found that 52.2 percent of ranked and all otherinstructional faculty In 60 state universities and land -grantcolleges were tenured. Furniss (1972) concluded that over50 percent of all instructional faculty are tenured at 46.1percent of the public universities, 61.3 percent of theprivate universities, 34.2 percent of the private four-yearcolleges, 37.1 percent of the public four year colleges, and43.1 percent of all 413 institutions which responded.

The current OIR tenure study of 62 graduate insti-tutions as shown in Table Three found that the number oftenured faculty increased by 9.7 percent between 1970.71and 1972.73. The number of tenured faculty increasedslightly less than twice as fast as the number of faculty eli-gible for tenure and about a third faster than the number of

130

all ranked institutional faculty over the same two.yearperiod. Table three also shows that over the two.yearperiod, in the upper two ranks, the number of ranked instructional personnel Increased by 10.3 percent, the numberof faculty eligible for tenure increased by 9.2 percent andthe number of tenured faculty increased by 10.6 percent.These figures indicate that faculty in the upper two rankshas increased approximately 10 permit ft:w all three cate-gories (ranked instructional faculty, faculty eligible fortenure, and tenured faculty) during the last two years.

Trow (1973) stated in his 1969 study of 27,191 facultythat tenured faculty were 51.0 percent of all ranked instruc-tional faculty (Includes ranked temporary faculty); 52,6percent of ranked instructional faculty eligible for tenure;and 49.9 percent of total instructional personnel (rankedand non-ranked, including lecturers, assistant instructors,research associates, and specialists). Compared with Trow,the current OIR study, as showin in Table Four, found thatfor 1970, tenured faculty were 55.1 percent of all ranked instructional faculty (includes ranked temporary faculty); 59.8percent of all instructional personnel (ranked and non-ranked). For 1972, tenured faculty were 56.6 percent of allranked instructional faculty, 62.3 percent of all ranked In-structional faculty eligible for tenure; and 47.2 percent ofall ranked and non-ranked instructional personnel.

The percentage differences between comparablecategories of the Trow study and the current OIR study are

Freeman arta liossmolor

Table 3

COMPARISON OF TENURED FACULTY BY RANKS DISTRIBUTIONFOR 62 GRADUATE UNIVERSITIES 1970-71, 1972-73

1970-1971

I II IIIInstructional Faculty Eligible Tenured

Faculty a for Tenure Faculty

N In 55,208%

N I.50,847%

N = 30,420%

Professor 29.6 30.6 50.9)54.6 )56.2 )86.9

Assoc. Professor 25.0 25.6 36.0

Ass't Professor 31.4 32.4 10.805.4 )43.8 )13.1

Instructor 14.0 11.4 2.3100.0 100.0 100,0

Professor

Assoc. Professor

Ass't Professor

I

N=58,967

30.6)56.4

25.8

31.603.6

1972.1973

IIINiest 33,375

51.3)87.6

36.3

9.8)12.4

IIN= 53,542

32.0)58.3

26.3

32.1)41.7

Instructor 12.0 2.6100.0

_AL100.0 100.0

II IIIPercent of Two-YearIncrease for all (3759) (2695) (2955)Ranks +6.7% +5.1% +9.7%

I II IllPercent of Two-Year increase In (3113) (2639) (2801)Upper Two Ranks + 10.3% + 9.2% +10.6%

a Included are temporary ranked faculty with pay.

131

Tenure

Table 4

PERCENTAGE OF TENURED FACULTY AT 62 GRADUATE INSTITUTIONSDEPENDING ON COMPUTATION BASE.

Total Total % of Ranked % of FacultyInstructional Instructional instructional Eligible forPersonnel Personnel' Who Faculty Who Tenure Who Are

Are Tenured Are Tenured Tenured

1972 112,693 47.2 56.6 62.3

1970 66,010 46.1 55.1 59.8

Includes assistant Instructors, lecturers, research associates, specialists,In addition to all ranked instructional personnel,

attributed partly to the institutions included in eachsample. Trow's study concerned faculty from a crosssection of higher education while the MSU study focusedonly on graduate institutions. Nevertheless, both studiesindicate substantial differences depending on computationbase to calculate the percentage of tenured faculty. The evi-dence also indicates here that tenure percentages are in-creasing steadily regardless of which cor.putation base isused.

Even though the MSU study has derived its trenddata on two years which are relatively close, one canassume that these trends tend to provide an overall pictureof the tenure situation, From the trends shown in thecurrent research and pointed out in previous research, thenumber of tenured faculty is increasing at a rate (.5 to .7percent per year) which could easily lead to disastrousconsequences in forthcoming years,

Current National Issues Affecting TenureIn view of the current tenure trend, what are some of

the current national issues which make a review of currenttenure personnel policies necessary? First, the end of themilitary draft, higher fees and tuitions, and a phenomenoncalled "stopping out by the Carnegie Commission arecausing a dramatic slowdown in enrollment growth at thenations' colleges and universities. The 10 percent growthrates of the 1960s have suddenly shifted to a more normal 23 percent level, and probably will remain at the lower rate

132

through the 1980's.Second, coupled with unparalleled enrollment

growth during the 1960s was the growth of institutionalbudgets. Now, in a time of general financial exigency,many institutions are finding themselves in a non-flexiblepredicament. Cheit, in his recent book, The New Depres-sion in Higher Education Two Years Later, claims thatthe 41 institutions he examined in 1970 have since then"achieved a stabilized financial situation and have gonefrom a financial condition of steady erosion to one offragile stability." The stabilized financial situation has beenaccomplished primarily through extraordinary cuts in ex-penditures that clearly cannot go on indefinitely. On theother hand, in the past year legislators, as well as adminis-trators, have found that budget reductions, programcutbacks or re-allocations became next to impossible be-cause of tenured faculty.

Third, severe pressures are being exerted on collegesand universities for increased accountability to a variety ofagencies and interests including the general public, legis-lators, governmental agencies, the courts, coordinating andgoverning boards, faculty, students, and other internal con-stituents. Mortimer (1972) cites three applications of theterra "accountability in higher education:" managerial ac-countability, accountability versus evaluation, and accountability versus responsibility. All three applications arerelevant to the proportion of tenured faculty in aninstitution,

Related also to the three dimensions of account-ability is the trend toward collective bargaining.? Willfaculty bargaining units begin trading off concepts of aca-demic freedom and tenure for pay increases or reduction inwork toads? Or will tenure bo bolstered as guaranteedsecurity through collective bargaining?

The fourth issue concerns longevity of currenttenured faculty. Miller (1970) claims that the averagetenured position extends over a 35year period, while theaverage nontenured faculty member stays only for sevenyears. Consequently, an institution has the choke betweenone tenured position or 10 nontenured positions over a 35-year period. Keast (1973) and Trow (1973) both claim thatwithin the last few years about 75 percent of all facultymembers were under 50 years old, and nearly two-thirds oftenured faculty were under 50. A tower average age distributton will result in fewer retirements for the next lenyears.

Closely related to tenured faculty longevity is theissue of ov Jr-supply. Faced with a surplus faculty and ascarcity of available positions, institutions aro experiencinga decreasing turnover rate because employed faculty areconstrained by considerably lessened mobility.

These issues, when placed in the concept of a nogrowth faculty situation, exacerbate the movement towardsrestricting tenure. Certainly such conditions as decliningrates of enrollment growth, dwindling financial resources,program reorganization, and fixed or reduced staffs shouldcause institutions to re-examine the current state of tenureaffairs.

institutional Self-ExaminationNational trend statistics are not as critical as indivi-

dual institutional situations, In reviewing your own situ-ation related to tenure levels, consider the following ques-tions. Assuming current personnel practices regardingtenure will continue unchanged, will your institution beable to deal with the following program factors:* future program shifts and new program development,* normal salary demands of faculty and staff,* extra demands of collective bargaining,* inflation and cost of living increases,* cutbacks in federal and state funds,* reversal of autofstate tuition clause,* highevest to retain tenure faculty (assuming that

the average salary of a nontenured faculty is abouthalf that of a tenured faculty member)?

In addition to determining an institutional tenurerate and rank distribution, each institution must determinerelative tenure rate and rank distribution for each collegeand department since percentages are not likely to beuniform throughout an institution. In fact, the situation Iscompounded if high tenure levels were found in decliningprograms, and not unlikely situation.

F'rooman and liossmolor

Solutions to High Tenure RatesUp to now, most institutions have not formally insti-

tuted any restrictions on the number or percentage oftenured faculty. In the OIR study of 130 graduate institu-tions. 13.4 percent of 82 institutions who responded indi-cated they were considering imposing a quota limitingtenure appointments. Furniss (1972) found that only 5.9percent of 413 institutions (all types) were actually limitingthe percentage of tenured faculty as of January, 1972.

Solutions applied elsewhere to the problems stem-ming from too many tenured faculty include tenure quotas,equalized rank distribution, early retirement benefits, termcontracts as opposed to a tenure system, more temporaryappointments, appointments of new faculty only to lowerranks, and a general slowdown of the rate of promotion. Ofthe procedures proposed for dealing with fixed commit-ment or tenure levels, some results are evident:a) Utilization of a package of benefits for early retirementdons not provide cost savings to the institution but is appro-priate for improving the distribution of faculty ago or rank.b) Appointment at a lower rank (instructor versus anis-tan% professor) and increased time between promotions iseffective in helping to sustain tenure levels or to reducethem. depending upon its method and rate of application.c) Stanford University (Hopkins, 1972) found, that if oneof its colleges made no additions to tenure white hiringnew personnel and allowing for terminations, the tenureproportion would drop from 71 percent back to 57 percentin a ten-year period, Michigan State University wouldexperience a similar kind of result.di Term contracts in lieu of tenure have not met withoverwhelming favor by those studying tenure and alterna-tive systems,

One key recommendation of the Commission onAcademic Tenure in Higher Education (1973) is to extendthe probationary period to at least five years before tenurecan be granted. Currently more than one-fourth of all insti-tutions have less than five-year probationary terms accord-ing to Furniss (1972). The Commission also recommendedthat tenure quotas be expressed and utilized as "ranges orlimits rather than as fixed percentages."

Concept of Flexible DollarsProbably more important than any of the mechan

isms and procedures Just given is the need for institutionsto consider those factors pertinent to the establishment ofwhat Furniss (1973) calls "steady-state" planning, that is,planning when an institution has stopped expanding,especially in enrollments. The suggestion is mado here thattenure percentages of committed positions, however com-puted, are not the critical factor in analyzing institutionalsituations. Rather, tenure commitment calculations shouldbe based on dollars rather than on positions. Table Fiveshows, that in a hypothetical situation where positions are

133

TENURE

converted to dollars, a different picture becomes evidentbetween dollar tenure commitment as opposed to positiontenure commitment because faculty in upper ranksconsume more dollars per position than lower ranks. Onthe dollar basis 64.60 percent of the total instructional costsis committed to tenured faculty. On the position basis 55.00percent of the total instructional faculty is committed totenured faculty. This illustrates the point that unless onetranslates obligations (tenure, tenure eligible, and jobsecurity) into dollars the appearances of low.tenured per.sonnel percentages can be dangerously misleading.

Derived from these comparisons is the concept offlexible dollars, which Is posed here as a better guide formanagement and planning In a steady-stato situation. Flex-

134

Ible dollars are those dollars that are not committed be.cause of tenure, job security, or the relatively fixed forms offinancial obligations. Flexible dollars are dollars (orfaculty) needed to accommodate student enrollment shiftsamong colleges and departments.

It is important for an institution to retain in eachbudget a sufficient amount of uncommitted dollars for non.tenured and nordenured.but.eligible faculty, so that post.tions and resources can be shifted to accommodate growthareas or provide for new academic programs, This ,nvolvesapplying a university SCH/FTEF average to all colleges anddepartments and then asking how many staff (FTEF)should be shifted (added or subtracted) from each college

meet instructional need. The shifts in faculty (FTEF) can

Table S

A HYPOTHETICAL EXAMPLE OF POSITIONS VERSUS DOLLARS.

FTE Positions

NTenuredFaculty

Percent of TotalFaculty with Tenure

Professors 25 25 25

Assoc. Professors 25 22 22

Ass't Professors 25 8 8

Instructors 25 0 0

100 55 55

FTE Dollars

Tenured Percent of DollarsN Dollars For Tenured Faculty

Professors $550,000 $550,000 33.3

Assoc. Professors 450,000 396,000 24.00

Ass't Professors 375,000 120,000 7.27

Instructors 275,000 el

$1,650,000 $1,066,000 6460

then be derived and converted to dollars. from which apercentage figure of the total Instructional budget isestimated for each year.

Hiring people at a lower rank and/or at a temporaryrank would improve the tenure ratio and would allow for acertain staff flexibility: however, the question of dollarflexibility rather than position flexibility may be unwiselyignored under such a procedure. Unless the number ofpeople on tenure or with tenure-like job security is re-duced, the institution's dollar flexibility has not improved.Holding constant the level of tenured people while increas-ing certain nontenured faculty makes an improved ratio butnot necessarily on improved dollar situation. For example,consider a reduction in programs cutting into temporary ornontenured personnel only, while leaving tenured staffalone. This is an attractive approach but generally an un-realistic one for long-range planning and management be-cause the long-range dollar flexibility Is further impaired.That is why some institutions have chosen to reduce entireprograms, thus reducing fixed obligations.

From the hypothetical situation in Table Five, 55 per-cent of tenured faculty translates Into a dollar commitmentof 60 to 70 percent of the operating budget, This table indi-cates that approximately 10 to 15 percent must be added toan institution's tenure personnel percentage figure to

Proomun and Itossmolor

obtain actual dollar commitment or the relative amount ofinflexible dollars, The question hero is at what level wouldthe percentage of positions become dangerously high? Theanswer, of course, varies according to institution. However,in general, a dangerous level is reached when actual dollarsto support committed tenured faculty hinders the institu-tion in coping with the financial constraints spoken ofearlier. A maximum level will also depend on salary levelsand the total position base for computing tenure, If oneincluded other personnel in the base, such as graduateassistants and assistant instructors, then the dollar factor isenlarged,

Because of factors such as future enrollment levels,state funding levels, degree of legal commitment to uni-versity personnel other than faculty, the changing nature offederal support, and external comparative data, one wouldconclude that for many institutions tenure levels and upperrank percentages are very close to their maximum. Cer-tainly, it seems unwise to allow commitments to growmuch buyond current levels if an Institution Is to be able tohandle any unknown future while retaining flexibility andInstitutional vitality. One the other hand, there is nocurrent evidence available which readily encourages insti-tutions to reduce their pattern of commitments,

I Eligible for tenure includes both tenured and nontenured faculty,

In early January. 1972, tenure information was requested from the 130 graduate institutions included in the publication by Kenneth D.Roose and Charles I. Andersen, A Rating of Graduate Programs, Washington: ACE, 1970. Usable information was received from 62universities,

As of January, 1973, there were bargaining units involving faculties at about 250 (or 8.0% ) of 2,800 colleges and universities in theUnited Stales. An estimated 75,000 faculty members were covered by these units. The New York Times of April U. 1973, reporting on a forthcoming book (Governance of I (igher Education: Six Priority Problems by the Carnegie Commission on Higher Education).

Blackburn. Robert . Tenure Aspects of Job Security on the Changing Campus. Research Monograph No. 19. Atlanta; SouthernRegional Education Board, 1972.

"College Enrollment Boom Peters Out The Ann Arbor News, April 11, 1973.

Commission on Academic Tenure in Higher Education, Faculty Tenure, San Francisco: lussey-Bass, 1973.

"Coping with Adversity Report on the. Economic Status of the Profession, 1971.72." AAUP Bulletin, 58 (2), (June 1972), pp. 178-243.

Poriiity Salaries in Lund-Grant Colleges and State Universities. Federal Security Agency, Office of Education, Circular No, 358, 195152.

Furniss, W. Todd. Facility Tenure and Contract Systems, Current Practice, American Council on Education Special Report. 1972.

Steady-State Staffing in Tenure-Grunting Institutions, and Belated Papers. Washington, D,G; American Council on Education, 1973.

Hopkins. S,P. David, An Early Retirement Program for the Stanford Vacuity; Report and Recommendations, Report No. 72.1. Stanford:Office of the Vice President and Provost, Stanford University, 1972,

The influence of Appointment, Promotion, and Retirement Policies on Facility Rank Distributions. Report No. 72-2, Stanford: Office ofthe Vice President and Provost, Stanford tiniversily, 1972,

Kees!. William R. The Commission on Academic Tenure in Higher Education: A Preview of the Report." presented at the annualmeeting of The Association of American Colleges, San Francisco, Ian uary 15, 1973.

Miller, John Perry. "Tenure: Bulwark of Academic Freedom and Brake on Change." Educational Record. 51 (3), 1970, pp, 241.245.

135

Tenure

Merlimer, Kenneth P. Accountability in Higher Education. Washington, D,C,: American Association fur higher Education, prepared byERIC Clearinghouse on Higher Education, 1972.

"The New Depresselon in Iligher Educalion Two Years Later 'From a Slate of Steady Erosion to One of Fragile Stabi Pty.TheChronicle of ffigher f.,'clucation. VII (28), (April 16, 1973), pp, 4.5, quoting Earl F. Chen, The New Depression in I figher Education; Two YearsUtter. Carnegie Commission, 1973.

Pfnisier, Alin 0. "Promotion and Tenure Policies in Undergraduate Colleges." North Central Association Quarterly, XXXII (3),(January 1950), 260.275.

"Retain Tenure Hut Ration II, Panel Advises." The Chronicle of Higher Education, Vil (January 22, 1973). pp. I. 5,

Row, Kenneth I)., and Andersen, Charles I. A Ruling of Graduate Progrums. Washington, D.C. American Council en Education, 1970,

Salaries Paid and Salary Practices In Universities, Colleges, 1961-62. Higher Education Series, Research Report 1962452, ResearchDivision, National Education Association, February, 1962,

Shaw, H.N. Academic Tenure In American Higher Education,Chicago: Adams Press. 1971,

Ha rdwell I.., and Associates, The Tenure Debate. San Francisco: JosseyHass, 1973,

"Summary of 'tenure Panel's 47 Recommendations." The Chronicle of Higher Education, VII (25), (March 26. 1973), p. 6.

Trow, Marlin. "The Distribution of Academic Tenure in American Higher Education," in Smith, Ha rdwell and Associates, TheTenure Debate. San Francisco: losseyllass, 1971

136

THE TEACHING EFFECTIVENESS OF AN ALTERNATIVE TEACHING FACILITY

To an increasingly greater extent we findourselves being arranged by impersonal environ-ments in lecture halls, airports, waiting rooms, andlobbies . , The straight-row arrangement ofmost classrooms has been taken for granted for toolong. The typical long narrow shape of a classroomresulted from a desire to get light across the room.The front of each room was determined bywindow location, since pupils had to be seated sothat window light came from the left shoulder.However, new developments in lighting acoustics,ventilation, and fireproofing have rendered invalidmany of the arguments for the boxlike room withstraight rows. (Sommer, 1967, p. 151)

Many colleges and universities have sought toovercome the rigidity of the straightrow" classroom insome of their teaching modules, but few have taken thetrouble to see if these new teaching facilities have had anyeffect on the learning of students.

In 1969, the University of Alberta General FacultiesCouncil created a Committee to Investigate Teaching. Thiscommittee addressed itself to all facets of teaching, fromphysical models to pedagogical techniques. One of theprojects undertaken by this committee was the designing ofan alternative teaching facility.

Three objectives were sought: first, to provide themaximum amount of versatility in intramural design andmanipulation; second, to equip it with more visual andtactual stimulants than is'normally found in conventionalclassrooms; and third, to study the effect of a lounge-typeclassroom on teaching and learning.

Maximum internal flexibility was achieved in thefollowing manner. Seats were portable half-hexagonalboxes which, when stood on edge, could serve as eithersmall work benches or waist-high partitions. Movablepanels were suspended from an ellipitical bar on the ceil-ing. These panels could be strung out the full length of theellipse, forming an egg shaped "womb", or could be com-pressed into a few feet, leaving the room open. Each panelcould rotate 360 degrees, and could be held in place bymeans of a locked set of casters. Four banks of electricallycharged rails housed incandescent pin lights. Each bankwas controlled by a dimmer switch. Each pin light fixture

Peter Horowitz and David Otto, University of Alberta

could be: (1) pointed at any locus on a half-sphere, (2) dis-mounted and moved to another electrical bank in the room.or (3) turned off. Each could hold a color filter.

One could use these three elements, (seats. wallsand lights), to create whatever kind of learning site was de-sired. For example, a class could begin with all students ina single campfire type circle in the center of the room.When the need for buzz-groups arose, smaller groups couldmove to the corners, and the panels could function asscreens. Or the class may sit on one side of the room andview presentations by students on the other side wherehalfhexagonal boxes could serve as a work area and thepanels as backdrops.

The Committee's second objective was to rekindlethe learner's sensory awareness. Color and light were itsfirst concern. Color was everywhere: the ceiling was mid-night blue; the pin lights had red, blue, green and yellowfilters; the carpet was dark green; the boxes were coveredwith orange and dark green carpet. Only the walls wereleft white. Natural sunlight entered the room either throughthe clear window pane or through translucent plexiglasssections in the wall panels. Because the pin lights were di-rectional and operated by dimmer switches, certain areas ofthe room could be flooded with light while other areas ob-scured by shadow. Other optical stimuli were present inthe plexiglass, plywood and fiberboard patterns in thepanels. Tactile sensory input was provided by carpet on thefloor and on the boxliko seats.

Varying geometric forms were available. Squaresexisted in the 4' x 4' panel inserts. Seats were half-hexa-gonal. These accompanied the usual rectangular outlinesfound in most conventional classrooms. The panels'pseudo-wall structure removed the starkness of the right-angle corners. This elliptical shape developed a groupseating which encouraged more eye contact among studentswhile still maintaining Hall's (1966) Social Distance.

A room such as this represents the antithesis of theEssentialist philosophy of education, There are no un-cushioned, straight-back, wooden chairs: no controlled,structured surroundings and no "bare necessities" whichare the hallmarks of the Essentialists' frame of mind (SeeWingo, 1965). "Sesame Street" has demonstrated that onedoes not have to indulge in regimentation and strict self-

137

Teaching Effectiveness

discipline in order to learn. But the question remains dopeople learn more in a congenial environment?

Research Design

We hypothesized that learning occurs either equallywell, or even better in an Alternative Teaching Facility thanin the more traditional setting. Learning was operationallydefined as tho scholastic performance, as evaluated by acompetent grader, of two groups of regular undergraduatestudents' at the University of Alberta enrolled in two sec-tions of English 275 (Introduction to Prose). Scholastic per-formance was measured by the letter grades, A-B-C-D-F,given to the students' two term papers and a number grade(1.9,"9"is excellent) given on the final examination.

These two sections were taught by the same instruc-tor (Horowitz) in two different classrooms: 289 CentralAcademic Building (the experimental setting) and G-114,Biological Sciences Building (the controlled setting). In-sofar as was humanly possible, the lectures and discussionsin both classes were identical, Mr. Horowitz met with bothclasses at every assigned meeting, The syllabus readinglists, assignments, term paper topics and final examinationwere identical.

A Graduate Teaching Assistant in the Department ofEnglish, was engaged to grade tho papers before Mr, Horo-witz read them. Sho was not inbimed of the nature of thoexperiment nor was she given any opportunity to interactpersonally with the students, Furthermore, she had no ideawhich studotns belonged to which section, as the termpapers and final examination wero collected, when duo,into one bundle and delivered to the teaching assistant forgrading.

The Hanmon-Nelson test of mental ability was ad-ministered to the members of both sections during regularclass time in the term by Mr. Horowitz.

The Subjects

As this was a field study, the effect of many vari-ables could not be neutralized. We were able to estimatesome of their impact, however.

Results from the Hanmon-Nelson tost ins :ate thatthe two groups of students had essentially the same level ofability. Eighteen of the original 29 students attended classthe day the test was administered to the experimental groupand 18 of the 32 were in the control class session the dayfollowing.

Table I

MEANS AND T-TEST VALUES OF THE VERBAL, QUANTITATIVE AND TOTAL SCORES OF THE TWO GROUPSOF STUDENTS (N = 18 EACH) ON THE HANMON-NELSON TESTS OF MENTAL ABILITY, COLLEGE LEVEL.

r

138

Variable

VerbalQuantatative

Total

Exper'I

40.7821.72

62.50

Control T Value

38.22 -0.8822.39 0.35

60.61 -0.54

(Maximum scores: Verbal, 60; Quantitative, 40: Total, 100)

F Test differences of variances between these groups.

Variable Exper'i Control F Value

Verbal 32.30 118.07 3.655.

Quantative 34.09 29.66 1.149

Total 50.03 160.02 3.199"

sp < .01 ** p < .025

Tho control group had a much wider range of "Intelligence," as measured by this test, Overall, the averagescore on tho verbal portion of the lost was lower than thatof the experimental group but slightly higher on thequantatativo portion. The West failed to display any signifi.cant differences in the verbal or quantatativo recall of thosetwo groups, One can conclude that the eAperimental groupdid slightly better in the verbal area because it had a muchsmaller variance about the meen than did the control group.

The schools and colleges represented by these twogroups of students were unmatched, More students were

Horowitz and Otto

from the faculty of Arts and the school of Dental HygieneIn the experimental group, while more students from thofaculilea of Agriculture, Business & Commerce, Enginwring, Pharmacy, Physical Education and Science woro inthe control group. Tho bias, therefore, tends to lean towardsthe experimental group, as many more Arts students wereenrolled in the experimental group.

Tho initial male - female registration figures, Tableshowed a distinct partiality towards the experimentalgroup.

Table II

A 2 X 2 CHI-SQUARE MATRIX OFMALES AND FEMALES

MALES FEMALES TOTAL

Exper'l 8 21 29Control 19 11 30

TOTAL 27 32 59

x2 = 7.59 df = 1 p < .01

The high number of females in the experimentalgroup suggested a better grade achieving performance fromthat group. Carter. 1952; Edmiston, 1943: Marley. 1933; andSchmuck, 1965 have all demonstrated "kat female students,ceteris paribus. received higher grades than male students.

There was no significant difference in the classstanding (Freshman, Sophomore, etc.) of the two groups asonly one individual in each group was beyond Freshmanstanding. In terms of age, no differences were noted, Theaverage age of each section was eighteen.

The control group had the preferred time of day(Tuesdays and Thursdays from 1:00 to 12:30), while theexperimental group met at 8:00. Mondays, Wednesdays,and Fridays. The room used by the control group wassomewhat less desirable than even a typically traditionalclassroom in that it had no windows, a fairly low ceiling

and a propensity for echo.A survey of these intervening variables indicated

that the members of the experimental group would havehad a slight advantage over their counterparts in the controlgroup.

Results

As Table III clewly shows, there were no differencesbetween the grades received by the Experimental and Con-trot Groups for the two term papers and final examination,

The assumption that this alternative teaching facilityis as conducive to teaming as is a traditional classroom canbe supported, but the premise that the facility is more bene-ficial to learning (as herein defined) is not supported,

139

Teaching Effectiveness

TABLE IIIChi - Square Comparisons of the Grades Given

the Two Groups of Students Registeredin Two Sections of English 275

FIRST TERM PAPEREXPERIMENTAL CONTROL

GRADE GROUP GROUP TOTAL

A 2 1 3

B 9 11 20C 10 10 20D 5 5 10F 0 1 1

TOTAL 26 28 54

x2 = 1.46 df = 4 p < .80

SECOND TERM PAPEREXPERIMENT CONTROL

GRADE GROUP GROUP TOTAL

A 1 2 3

B 7 7 14C 11 10 21

D 3 3 6F 0 1 1

TOTAL 22 23 45

= 1.36 df = 4 p < .85

FINAL EXAMINATION

GRADEEXPERIMENTAL

GROUPCONTROL

GROUP TOTAL

9 0 1 1

8 4 4 8

7 8 7 15

6 11 5 16

5 4 3 7

4 0 3 3

3 0 1 1

TOTAL 27 24 51

= 7.31 df = 6 p < .30

140

Discussion

Wo do not feel that grades alone reflect all that hadoccurred in both classrooms. (See Maslow & Mintz, 1950and Mintz, 1950). A number of differences were noted inthe behavior of the sections. As the term wore on, we no-ticed that attendance was far better in the experimentalclass in spite of its being held at 8:00 a,m, The students inthe experimental group were more ready much earlier inthe term to participate in class discussions, and by lateOctober, were actively debating quite freely among them.selves and with the instructor, all of this with little urgingfrom the instructor, Students in the control group, on theother hand, needed prodding throughout the term and wereby and largo content to sit silently through most classes.

No observable differences existed between the stu-dents in either group on the dimension of receptivity to thematerial presented, but, as an example, half of one classsession in the experimental group was devoted to exam-ining the reasons why the class thought the book assignedfor that meeting was "utterly boring." Students in thecontrol group no doubt shared the same opinion of the as-signed reading, but no one thought to complain,

We noticed more informality and group cohesion inthe experimental group than in the control group. More ofthe students in the experimental group were addressingeach other by first names, and this occurred between stu-dents registered in different faculties. In contrast, fewstudents appeared to know, or even recognize, their class.mates in the control group.

Finally, Horowitz reports that visits during officehours were more numerous from students in the experi-mental group than from students in the control group, andthat during these visits, those from the experimental groupseemed more at ease than those from the control class.During the second term (English 275 was a first termcourse) several of the students from the experimentalsection continued to visit the instrurtor, who never sawanyone from the control group again.

Surprisingly, the cost of furnishing an AlternativeTeaching Facility is close to the cost of equiping a tradi-tional lecture room. Room 289, Central Academic Buildingdid cost more than it should, and more than the traditionalclassroom does, but this difference was chiefly due to mis-calculations in planning the pilot project and the need to

Horowitz and Otto

rewire the light fixtures, A second such teaching facilityhas been furnished in another building on campus. Whenthe cost of the traditional furniture was deducted from the$2,500 spent for reconversion, the next expense for thesecond room was approximately $1,000. Hero, again, $500was spent in removing flourescent fixtures and installingincandescent lamps, It would bo safe to estimate that thechief difference in furnishing a room such as 289, CentralAcademic Building in a building under construction wouldbe the price of a carpel.

Recapitulation

An Alternative Teaching Facility was created inorder to provide a more stimulating environment for learn-ing, The rotAts of the students' work in the cognitivedomain, when measured in terms of the conventionalsystem of grading, showed no noteworthy change. Informalinteraction, both between students and between instructorand student, elements of the affective domain, did, how-ever, increase.

One final comment. Within the last twenty years,psychologists such as Hebb have begun to explore theeffects of sensory deprivation on human beings. They havelearned that human subjects, when deprived of normalsensation for extended periods of time become bored, rest-less, mentally lethargic, and have reported an inability toengage in prolonged throught, (See Altman, 1971; Heron,1957; Hebb, et al 1954; and Vernon & Hoffman, 1958). Inan effort to provide the maximum amount of space percapital budget dollar expended, administrators and archi-tects have continued using rectangular box-like cubicleswith bolted down ranks of seats facing front, in an environ-ment of bland monochromaticlsm, washed in uniform(shadow diffusing) flouroscent illumination, and coveredwith sterile smooth-surfaced floor tile, (See Bechtel &Srivastaba, 1966; Black, 1950; Cooper & Zubek, 1958; Kyzar,1971; Maslow & Mintz, 1956; Mintz, 1956; Reichert, 1973;Sommer, 1969; and Wotton, 1970)

We thrust our students into the environment andexpect them to learn. In our opinion, creative, productivelearning is not compatable with this setting.

Alima, I, "Ecological Aspects of Interpersonal Functioning" in Esser, Ail. (ed) Behavior and Environment New York: Plenum Press,1971, pp. 291-306.

Argyle, M & Deon, I. "Eye Contact, Distance and Affiliation." Sudo/retry v. 28 (1965) pp. 289-304.

Bass B.M. & Klubeck. S. "Effects of Seating Arrangements on Leaderless Croup Discussion," Journal of Abnormal and SocialPsychology v. 47 (1952) pp. 724-727.

Bechtel, IL & Strivaslaba, It -Human Movement and Architectural Environment." Mileu V. 2 (1966) pp. 7.8.

141

Teaching Effectiveness

Bevelas, A. "Communication Patterns In TaskOriented Groups," Journal of the Acoustic°! Society of America v. 22 (1950) pp. 725.730,

Black, I.W, The Effect of Room Characteristics upon Vocal Intensity ,Ind Rate." journal of the Acoustical Society of America v, 22(1950) pp. 174.176.

Brodey, W. "Sound and Spaco" Journol of the American institute of Architects v. 42 (1964) pp. 58.60.

Carter, R.S. "How Invalid are Marks Assigned by Teacher?" Journal of Educational Psychology v, 43 (1952) pp. 218.228.

Cooper, R.M. & Zubek, J.P. "Effects of Enriched and Restricted Early Environments on the Learning Ability of Bright and Dull Rats"Conodion Journol of Psychology v. 12 (1958) pp. 159-164.

Edmiston, R.W. "Do Teachers Show Partiality Toward Boys or Girls?" Peabody liournal of Educotion v. 20 (1943) pp. 234-238.

Filipe, N "Interpersonal Distance and Small Group Interaction" Cornell Journol of Social Helotions v. 1 (1988) pp, 59-64.

Frye, R& Standhardt, F. "See MoreHear MoreLearn More in Windowless Rooms" Educotional Screen and AudioVisuol Guide v. 40(1961) pp. 274-277.

hall E.T. Tho Hidden Dimension New York: Doubleday & Co. inc., 1966.

Baru AP, & Bales, R.F. "Seating Position and Small Group Interaction" Soclometry v. 24 (1963) pp. 480-486.

Harmon, D.B. "Lighting and Child Development" Illuminoting Engineering v, 40 (1945) pp. 199-228.

Hearn, G. "Leadership and the Spatial Factor in Small Groups" Journol of Abnormol ond Sociol Psychology v, 54 (1957) pp. 269-272.

Heron, W. "The Pathology of Boredom" Scientific Americon v, 198 (1957) pp. 52-58.

Horowitz, M.J. "Human Spatial Behavior" Americon Jo urnot of Psychotherapy v. 19 (1965) pp. 20-28.

Howells. L.T. & Becker, S.W. "Seating Arrangment and Leadership Emergence" Journal of Abnormal ond Sociol Psychology v. 64(1982) pp. 148.150.

Kyzar, B.L. "Comparison of Instructional Practices in Classrooms of Different Design. Final Report" Natchitoches. Louisiana, The SchoolPlanning Laboratory of Northwestern Slate College, 1971 (mimeograph), ERIC report No, ED 048 559.

Lewis, C.F, "Equipping the Classroom as a Learning and 'reaching Laboratory" Americon Schoolboord Journol v. 90 (1940) p. 29.

Maney. C.E. "Sex Was in College Marking" Journol of Higher Educotion v. 4 (1933) pp. 29.31.

Maslow, AIL & Mintz, N.L. "Effects of Esthetic Surroundings: I Initial Effects of Three Esthetic Conditions upon Perceiving 'Energy'and 'Well Being' in Paces" journot of Psychology v. 41 (1956) pp. 247.254.

McBride, G. A General Theory of Socioi Crb ^nizution ond Behoviour St. Lucia: University of Queensland Press, 1964.

Mintz, N.L. "Effects of Esthetic Surroundings: II. Prolonged and Repeated Experience in a 'beautiful' and an 'Ugly' Room" bum, ofPsychology v. 41 (1956) pp. 459-468.

Reichert, J.F. "Why Not Color Your Campus Creatively?" The Chronicle of Higher Educolion v. 7 (March 19, 1973) p. 12.

Schmuch, R. & Van Esmond, E. "Sex Differences in the Relationship of Interpersonal Perceptions of Academic Performance"Psychology in the Schools. v. 2 (1965) pp. 32-40.

Sanders, D.C. ''Innovations in Elementary School Classroom Seating" Bureau of Laboratory Schools Publication No. la Austin, TexasUniversity of Texas. 1958.

Sommer, R. "The Distance for Comfortable Conversation: aurther Study" Soclometry v. 25 (1962) pp. 111.116.

Sommer It "Leadership and Group Geography" Soclometry v. 24 (1961) pp. 99.110,

Sommer, Et. Personol Spoce, the Behoviorol Bosis of Design Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1969.

Sommer, R' "Small Group Ecology" Psychotogicol Bulletin v. 87 (1967) pp. 145-152,

Sommer, R' "Studies in Personal Space" Soclometry v. 22 (1959) pp. 247-200.

Sommer, It & Whitney, G. "Design for Friendship" Conodlan Architect v. 6 (1981) pp. 59-81.

142

Horowitz and Otto

Slelnzor, B. "The Spatial Factor In Paco to Face Discussion Groups" Journal of Abnormal and Social Psychology v, 45 (1950) pp. 552555.

Studer, R. & Sled, a "Architectural Programming, Environmental Design. and Human Behavior" Journal of Social Imes v. 22 (1966) pp.127436

Vernon, J. and Hoffman, J. "Effect of Sensory Deprivation on Learning Rate In Human Beings" Science v, 123 (1956) pp. 1074.1075.

Wing°, C.M. The Philosophy of American Education Now York: D.C. Heath and Company, 1965,Wotton, E. "Environment and Motivation A Close Relationship" Transcript of a Conference held In the Public Archloves and National

Library Auditorium, October 8th, 1970 (mimeograph).

"ION

143

RESEARCH REPUTATION AND TEACHING QUALITY IN UNIVERSITY DEPARTMENTS

Judy Richardson, Terry Each) and John McMillin, University of Washington

Studies of the relationship between research andleaching h t9 provided little support for either side in theongoing debate of the relationship between teaching andresearch. Voeks (1962), Stallings and Singhal (1970), andHayes (1971) all found no significant relationship betweenpublication rate and student ratings. Only Maier (1968)found a positive relationship between research andteaching: at Tufts University, faculty who had receivedresearch grants received higher student ratings than thosewho had not. And the most recent review of existingresearch and opinion clearly presents the need for furtherinquiry (Page, 1972).

Measure of Research ReputationThis study differs from the earlier research mainly in

that we examined the research productivity and teachingquality of departments, rather than of individual facultymembers. The measure of departmental research reputationwas from the most recent American Council of EducationRating of Graduate Programs (Roose and Anderson, 1970).Since a department's A.C.E. rating is based upon a nationwide survey of scholars in that field, it is essentially a mea-sure of the quality and the quantity of research conductedby the faculty of the department.

The darn provided in the A.C.E. report enabled us todevise two systems for ranking University of Washingtondepartments according to research reputation. In the first(or raw score) system, the departments were ranked accord.ing to the percentage of respondents who ranked the de-partment's faculty as distinguished and strong. The second(or percentile) ranking system was based on each depart.ment's relative rank among the other departments of thatdiscipline across the country. For each of the seventeenUniversity of Washington departments, we calculated a national percentile score, and then the departments wereranked on the basis of this score. (See Table 1.)

The differonce between the two ranking systems isthat in the raw sec re system, University of Washington de-partmenta were compared with each other directly, whereasin the percentile system, the departments were comparedfirst with their field nationally, and then with the other

144

University of Washington departments. 'rho purpose of thepercentile ranking system was to reduce possible dif-ferences in the generosity of raters among the various disci-plines.

We also gathered departmental operating data, whichserved as additional indicators of departmental research ac.tivity. This included (among other data) the percentage offaculty time spent on research.

Measure of Teaching QualityTo measure teaching quality, we designed a new stu

dent rating form. Since we wanted to insure maximumagreement on the meaning of a response, we employed aformat of scale descriptors to aid the respondent,

The choice of items was determined by our desire tokeep the instrument short while accommodating two kindsof scales: those that measured the incorporation of researchinto teaching, and those that measured the generally.ac.cepted attributes of good teaching. The research-orientedscales were: Knowledgu of Subject, Currentness ofMaterial, and Use of Own Research, All of the researchscales and most of the general scales were chosen fromthose that discriminate well between the best and worstteachers (Hildebrand, et al., W71). (See Illustration.)

Sample

From a list of individual programs of studies pro-vided by the Registrar, 1106 students taking courses in theseventeen departments were randomly selected to receivethe rating form. Returns were received from 58% of thesample; therefore, our analysis was based upon 643 studentratings. To protect tho anonymity of individual instructors,the returned ratings forms were identified only by depart-ment and level of instruction.

ResultsUsing first the raw score ranking system, rankorder

correlations were calculated' between A,C.E, ratings andmean student ratings on all items, and between A.C.E.ratings and mean student ratings on the research itemsalone. The correlations were .16 (Xi -a) and -.07 (ER)

Richardson, gado and McMillin

TABLE I

RAW-SCORE, PERCENTILE, AND STUDENTRATING RANKSBY DEPARTMENT

GHE

R

DDL

1

0WE

R

Department

ACE RAW SCORE RANK

R

1

2

3

4

5

6

7

8

9

10

11

11

11

11

11

11

11

ACE PERCENTILE RANI(STUDENT RATINGS

RANK

Percentage Who RatedDepartment Distinguished

and Strong xir x % Ile R

2

8

2

1

2

16

2

6

8

14

7

15

10

17

13

10

10

7 X1.8 R

5 3

8 8

15 16

7 12

,'N 6 7

3 4

11 14

14 13

10 9

1 1

16 15

2 2

17 17

13 10

4 5

12 6

9 11

Geography

German

Mathematics

English

Sociology

Anthropology

Physics

Chemistry

Psychology

French

Economics

Geology

MechanicalEngineering

Music

Philosophy

PoliticalScience

Spanish

58

49

40

39

36

29

25

22

17

16

6/34

12/48

18/102

12/92

13/73

16/42

20/113

24/125

27/110

19/63

19/91

22/69

20/71

18/43

19/65

21/74

18/65

82

75

82

87

82

62

82

81

75

71

80

69

73

59

72

73

73

These departments are in a score range placing them below French, but no percentageswere given. They are listed alphabetically.

145

RESEARCH AND TEACHING

illustration

The following scale consists of qualities which instructors may possess to varying degrees.Please rate the Instructor of the course listed above on each the the qualities. Each qualityis divided into three sections, and each section into three degrees, numbered accordingly

from 1 to 9. in rating, circle the number which best describes your instructor.

1 2 3 4 s 6 7 a 9

ENTHUSIASM Is definitely enthusiastic about h at times enthusiastic about subject Seems lo teach course without enthusiasm.FOR SUBJECT matter of course, mater,

I 3 3 4 $ b 7 I 9

KNOWLEDGE Definitely an expert; well-read and In good command of suSlect matter. Unsure of subject matter, not welireld.Of MIKE' experienced.

I 2 3 4 S a 7 8 9

TOLERANCE Encourages and accepts widely differing Accepts views somewhat differing Isom his Discourages differing points of view;points of tole*. own, intolerant.

3

CURRENTNESS Keeps course up to dates presents resultsOf MATERIAL of current revival., or sckriarship;

Introduces still debateable Issues.

PRESENTATION Stimulating and exciting.

1 2

AVAILAIIILITYTO STUDENTS

USE Of OWNRESEARCH

4

Occasionally mentions current or recentresearch in the field.

Moderately Interesting,

3 4

Is often available and easily approached forformal or informal contact In addition toposted office hours.

2

6

9

Course preparation could have been madeyears ago.

Puts you to sleep.

1 9

5 6 Z. I 9

Reasonably friendly; can be approached afterdais and during posted office hours,

3 4

Consistently incorporates OW/1 research Intocourse. course.

S 6

Resists student contact, rarely asailabteafter class; rarely keeps, or has noregular office hours.

7 a 9

Occasionally incorporates own research Into Little or no evidence of Involvement Inreseirch.

2 3

PREPARATION Shows definite evidence of carefulpreparation.

4

Shows some preparation.

both insignificant. Having found no overall correlation be-tween research reputation and teaching quality, we thencompared departments by field of study. For the socialsciences alone, we found a slight positive relationship be-tween A.C.E. ratings and student ratings on the research-oriented items. This suggested that for the social sciences,at least, departments with greater research reputation mayincorporate research into instruction more than those oflesser reputation.

We next calculated the rank-order correlation be-tween A.C.E. ratings and student ratings. using the per-centile ranking system. Th:s time, we found slight negativecorrelations: -.37 (Xi) and -.58 (Ks). The negative corre-lation between national percentile ranking and studentratings on the research-oriented items was statisticallysignificant at the .05 level. This suggested that those de-partments which had the highest relative rank in the

140

S 6 1 a 9

Not well prepared.

country were perceived by students as incorporating re-search into teaching to a lesser extent than those with lowerrelative ranking.

When the data were analyzed in an attempt to account for this finding, we discovered that the major contri-butors to this negative correlation were the physicalsciences. For example, Math and Physics were tied forsecond in percentile rank, yet received nearly the loweststudent ratings on the ressarch-oriented questions.

This finding, together with the slight positive re-lationship found for the social sciences in the raw scoreranking system, led us to conclude that there might be sig-nificant differences in the research - teaching relationshipamong the various fields of study as well as among depart-ments within fields. It is also possible that these relation.ships are obscured when all departments are combined to-gether for analysis.

Mean Student RatingsAll Items (X1.8)

High 2.0

3.0

Low 4.0

Figure 1

I

r = -.49

I

Richardson, Eado and McMillin

II

lirm.......e./...rosw..............r.......alhorovi

8 12 18 24

Percentage of Faculty Time Spent on DepartmentalResearch

High 2.0

Mean Student Ratings 3.0

ResearchRelated Items (RR)

I

4a

Low 4.0 ' r = -.35 5

6 12 18 24

Percentage of Faculty Time Spent on DepartmentalResearch

147

RESEARCH AND TBACHING

The differences in the research-teaching relationshipamong various fields of study appear to be the result ofthree major subject matter differences, First, subjectmatters differ In characteristic typo of research (rangingfrom experimental to general scholarship) and In the extentto which incorporation of research into classroom instruc-tion can be recognized and evaluated. For example, In thosocial sciences, research is generally experimental or des-criptive, Methodology is stressed and studies are always re-ferred to by author, so it is highly unlikely that a professorcould refer to his own work without a student's beingaware of It. In literature or philosophy, however, studentsare unlikely to know whether an idea presented by a pro-fessor came off the top of his head, from notes he took in agraduate seminar, or was the result of scholarly research.

Second, subject matters differ in the size of the gapbetween undergraduate and graduate level of understand-ing, For example, current research in psychology orsociology can usually be understood by students even inIntroductory classes, whereas it would be very difficult toexplain the latest research in physics to beginning physicsstudents,

Related to this are subject matter differences in thesize of the gap between the educational needs of thegeneralist and specialist. Here again, the gap in the socialsciences appears to be smaller than that In the physical sci-ences, For example, the latest research on cognitive dis-sonance may help the freshman In a survey course to betterunderstand his behavior, whereas the latest research onneutrons is of doubtful relevance to the nursing studenttaking a "service course" in chemistry.

As a result of these findings. we feel that futurestudies of the relationship between research and teachingneed to be undertaken at the subject matter level, byspecialists In that field working with specialists in institu-tional research. Such studies might compare departmentsacross several universities. One of the less complexvariables to examine is time. For example, in this study we

examined by department the relationship between qualityof teaching and the percentage of time spent by teachingfaculty In research, A negative correlation was found. Asthe regression lines In Figure I Indicate, this relationshipwas almost Identical for teaching quality as measured bytho means of all items(Xt s)and that measured by the meansof the three research-related items(Xs).

ConclusionIn closing we emphasize the tentative nature of our

findings and urge that more delicate and sophisticatedanalyses be undertaken, Some of the factors to be con-sidered have been suggested by this study. However,several additional considerations need to be taken into ac-count in future studies of research and teaching. The firstof these is the difference in kind of research that may beundertaken. Basic research, applied research, and researchon educational problems may all have different effects onteaching.

A second consideration is that In order for researchto have to beneficial influence or. .teaching, it is not neces-sary that either research quality or quantity be correlatedwith teaching quality, Instead, it Is only necessary that theteaching quality of those engaged in both teaching and re-search be higher than it would be, were they only teaching.Thus, rather than comparing teaching quality of faculty Inthe dual role, future studies might compare the teaching ofuniversity teachers with that of full -time teachers at otherInstitutions of higher education.

Finally, given the subtle interdependencies of re-search and teaching and the adherents to doctInaire posi-tions for and against the present system of dual roles, greatcare is required in the presentation, interpretation, and dis-cussion of the results of future studies, Offending thesensitiveness of defenders of the status quo might provokeunfortunate resistence to any improvements that might beindicated as a result of such studies.

A note of appreciation is extended to professor Robert Cope who wrote the U.S.O.E, grant proposal and assisted in the analysis.

Bresier, lack B. "Teaching Effectiveness and Government Awards," Scinece, 160 3824. April 12, 1968, pp. 164.167.

Cope, Robert G., John McMillin, and Judy Richardson. A Study of the Relationship Between Quality Instruction as Perceived byStudents and Research Praductivity in Academic Departments. Final Research Report to U.S.O.E. November, 1972.

Hayes, John R. "Research, Teaching, and Faculty Fate." Science, 172: 3980, April 16, 1971, pp 227.230,

Hildebrand, Milton. Robert Wilson. and Evelyn Dienst, Evaluating University Teaching. 13erkeley: University of California Center forResearch and Development in Higher Education, 1971,

Page. Cohn flood. 'Teaching and research-happy symbiosis or hidden Warfare?" Universities Quarterly (Britain), Winter 1972, pp. 102118.

148

Richardson, Eado and McMillin

Roose, K.D, and C.J. Anderson, A Rating of Graduate Programs. Washington: American Council on Education, 1970.

Stallings. William M. and Sushi la Singhal. ''Some observations on ihe Relationships Between Research Productivity and StudentEvaluations or Courses and Teaching," Thy American Sociologist, 5:2, May 1970, pp. 141143.

Voe lcs, Virginia. "Publications and Teaching Effectiveness: A Search for Some Relationship," fournal of Higher Education, 33, 1902. p.212.

N ON

149

SEXUAL BIAS ANALYSIS IN HIGHER EDUCATION;AN APPRAISAL OF METHODOLOGY USEFUL TO INSTITUTIONAL RESEARCHERS

Michael J. LaBay and Randolph N. Poster, Youngstown State University

OverviewWe are now well acquainted with the legislative his.

tory of the "affirmative action" doctrine of 1967. IntroducedIn ridicule by prominent Dixicrats opposed to the 1964Civil Rights Act, the doctrine took the form of ExecutiveOrder Number 11375 and stated in part that:

. (A federal) contractor will not discriminateagainst any employee or applicant because of race,color, religion, sex, or national origin. Thecontractor will take affirmative action to ensurethat employees are treated during employment,without regard to their race. thlor, religion, sex ornational origin (Seabury, 1972, p. 39)

The Dixicrats who introduced the word "sex" to thatCivil Rights Act as an amendment never really expected theabove executive order to be issued. As one contemporarywriter has stated, the afterthought amendment was offeredprimarily "to rouse . Northern masculine ire against thewhole bill" (Seabury, 1972, p. 38).

A short time after the Executive Order was signed,however, the Department of Labor quickly offered furtherguidelines under Order Number Four which read, in part:

... an acceptable affirmative action program mustincude an analysis of areas within which the con-tractor is deficient in the utilization of minoritygroups and women" (Seabury, 1972, p. 39).

Thus, the above guideline became the first national direc-tive issued by a federal agency to universities concerningsexual discrimination allegation procedures on campus,

It took little time for the Labor Department's direc-tive to be cited by the Department of Health, Education andWelfare (HEW) as applicable to university hiring and re-muneration practices. Executive Order 11375 quickly af-fected both public and private institutions of higher educa-tion. Columbia Unkursity and The University of Michiganobtained the unenviable distinction of being the first insti-tutions of higher education where unacceptable affirmativeaction plans were defined by HEW authorities (Zwerdling,1973).

Most institutional researchers realized that, long be-fore the mandate of "affirmative action," universities haveroutinely reviewed their own hiring and remunerationpractices. It is simply not good administrative practice to

150

omit such analysis. Both faculty rapport and fiscal account-ing practices rely heavily upon information securedthrough such studies.

With the HEW mandate, however, these routine in-stitutional analyses emerged as a public as well as a privateconcern. Using Executive Order 11246, the Women's EquityAction League (WEAL) recently asked HEW to investigatealleged sexual discrimination throughout the State University Systems of Michigan, New York, California, Florida,and New Jersey. Under fire from the Contract Compliancedivision of HEW, the University of Michigan has recentlywritten the nation's first Affirmativa,adinn plan of equalsexual employment opportunities.

Other instances of anti-discrimination litigation andsubsequent policy reorientation by colleges and univer-sities could be cited. It is enough at this point to simplystate that, regardless of how today's institutional researcherperceives his function, he is fast becoming the dejuro anti-discrimination fact-finder for his university.

Statement of PurposeSomething is wrong when a sexual bias allegation is

easier to specify than is a related documentation method-ology. Yet this appears to be that state of affairs in many oftoday's universities. 'Phis paper will attempt to show that itis only when researchers have not had the opportunity toeither define relevant stratifying variables on which hiringand remuneration practices are based, or when they attempt to use statistical analyses inappropriate to the deci-sions to be made that documentation becomes less efficientthan the precipitating allegation.

Let us begin with an examination of existing quanti-tative methods applicable to discrimination studies, followwith recommended investigative procedures deemed usefulto institutional researchers charged with related analyseson their campus, and end with an example of one serviceremuneration study successfully conducted at YoungstownState University (Youngstown, Ohio) during the Spring of1971

Review of Available MethodologyA reader performing a cursory review of existing

literature will quickly find that few authors have written

specifically on the topic of higher education serviceremuneration policy and related post hoc analyses. Indeed,at the date of this writing, HEW itself has failed to pre-scribe quantitative methodology commensurate with theiraffirmative action guidelines.

Today's administrative researcher must rely onspecific methods at once unique to his institution and ac-ceptable to affirmative action officials outside his organi-zation. Fortunately, an increased coherence of researchfocus and procedures are now beginning to be reported inthe professional journals. Many reports appear, at firstglance, to be so contextually distinct from each other thatthe degree of external applicability seems low. Closer in-spection, however, reveals that contemporary analysisprocedures have evolved within a three-stage refinement oftechnique. Let us look briefly at these evolutionary trends.

The quantitative theory of faculty selection and re-muneration procedure analysis seems to have been refinedthrough three stages. Initially only an elementary quotamodel was used. Subsequent refinements to date have in-cluded a second regression model and a third Bayesian ordecision theory model.

The quota model applicable in selection analysiswas understandibly refined during the mid-to-late sixtieswhen American social conscience was most critical of mi-nority oppression. Minority members were pictured asrunning a foot race with the majority. Track officials,noting that to stop the race, unshackle the American Blacks,then allow the runners to continue would still produce anunfair race providing the shackled runner with lime to"catch up" before continuance of the contest,

With the acceptance of this analogy the quota modelwas tentatively accepted. Given this model it was a rela-tively simple matter for institutional researchers to definethe presence or absence of selective racial or sexual bias bymeasuring the percent of qualifiable minority applicantswho were hired and promoted within their institution. ifthe percentage approximated the body count of minority inthe population of all potential qualified applicants,minority repression allegations could assuredly be dis-missed.

It took little time for officials to discern the short-comings of the quota model. ft failed to define adequate ex-pertise levels of potential candidates, and "tooling-up" theminority using majority guidelines was not as easy as somehad anticipated. Articles questioning and belittling quotamodel use began to appear. On November 6, 1972, a Berke-ley Department Chairman was quoted in a national newsweekly as stating: "We really looked for a qualified woman,but we just couldn't find one" (Newsweek, November 6,1972, p. 114). At approximately the same time, GovernorMilton Shapp (Pennsylvania) directed his state agencymanagers to recruit, train and promote as many women and

LaBay and Poster

other minority members as they could that year, but not tointerpret his directive as representing a quota system. Ob-viously, even politicians were realizing that quota alonewas not good enough.

Other social critics voiced displeasure in quotas withselective training riders, intellectuals of the Jewish faithemphasized that, since the Carnegie Commission on HigherEducation had determined that Jews comprise nine percentof the faculty in the nation's population (as contrasted withthree percent of the nation's population), the affirmativeaction doctrine based on quotas and selective training mandated that Jews stop running their own footraces until themore recently unshakled acquire identical training and pro-portional membership. With increasing frequency othercritics voiced their opposition, and the quota model wasquietly dismissed.

In its place emerged a second regression model, astypified by the studies of Loeb and Ferber (1971) and Astinand Bayer (1972). Multiple linear regression seemed, at firstglance, to be better suited to faculty selection and remuner-ation bias analysis than quota analysis. It necessarily in-cluded indices of expertise which the simpler quota modelfailed to define, and it directly related the indices to ac-cepted competency criteria.

The multiple regression equations were based uponan extension of bivariato linear analysis which producesprediction equations of the form:

AY = V + b(XR)

AWhere Y = predicted salary of candidate Yi for his ex-

pertise level

b = regression coefficient of Y on X produced bydifferentiating with respect to X and settingIt equal to zero.

= arithmetic mean of the X distribution

In multiple linear regression, more than one pre-dictor (X) may be used to estimate the value of the criterion(Y). If Yi denotes salary remuneration of t individuals and

represents a value of the ith expertise predictor (e.g.,years of service in the profession, terminal degree, numberof publications, etc.) for the tth individual faculty member,the partial linear regression equation useful for obtainingthe weighted sum of is:

whereAYs =

AYi = au + biXit + bkko

is a predicted placement within a rank ordercandidate list or an existing faculty mem-ber's predicted salary

151

Sexual Bias

In the above equation, an is a constant, and the b's(In.,.bk) are defined ar: partial regression coefficients andaro also constants. If the b's aro reduced to standard scoresso that comparisons across expertise areas can bo made, theresultant standard measure (13's) aro known as standardpartial regression coefficients. They are useful to adminisIrak° researchers in that they are independent of originalpredictor units and may be used as an index of comparativeweights each expertise predictor contributes to the predic-tive power of the equation. When using the /3's property,the researcher defines the variance of a composite of nweighted predictors as

ay.', me + 2E rij 131 131

where asq = variance of a sum of n weighted predictors

p = standard partial regression coefficient of thepredictor

riJ = correlation between Xi and any other var-where fable Xi where j>

I = standard partial regression coefficient ofthe predictor, with j>

Usually a multiple regression analysis produces amultiple coefficient of correlation (R12 I), which is derivedin the analysis simultaneously with (3 I and bi generations.Like all productmoment correlation coefficients, R indi-cates .the extent of relationship between two interval orbetter measures linearly related, In salary studies. it be-comes the correlation between salary and the total collec-tive of stratified predictor variables consisting of theweighted sum of score generated in the /3 and b determina-tions, In geometric terms. R is the slope of the teL3t squaresline of best fit connecting these two variables of sup-posedly equal variability. The obtained R Is always a maxi-mum value when the regression weights have been cor-rectly defined. The terminal product of multiple linear re-gression analysis usually consists of a multiple correlationcoefficient and prediction equation such as:

AY = 9600 + 200X1 + 10X2 + 10X3 + 500X4

In this ficticious example, if Y represented averagesalary of a predefined population and X1..,X4 representedpredictors properly represented by an interval scale and ofapproximately equal variability, the researcher couldproperly state that every unit increase in XI (e.g., number ofpublications) is associated with a salary increase of $200.

As in the case of bivariate linear regression, theabove prediction equation has an associated standard error

152

of estimate and related variance partitioning proceduresuseful in establishing confidence intervals and checkingthe assumption that the relation invoqligated is truly linear.

As noted earlier, studies by Loeb and Ferber (1971),and Astin and Bayer (1972) used multiple regressionanalysis to investigate allegations of sexual discrimination.Loeb and Ferber's techniquo involved a proportioning ofvariance prior to regression analysis, while Astin and Bayeremployed a stepwise multiple regression analysis directlyon a multitude of expertise variables. Both of these studiesare worth reading. Care, however, should be taken by thoreader in interpreting some results. Researchers acquaintedwith basic linear regression using least squares criterionand the general linear hypothesis techniques developed byBottenberg and Ward (1963), will quickly see unwarranteddata assumptions in both publications.

Assuming for the moment that their assumptionswere justified, Loeb and Ferber reasoned that sexual dis-crimination existed if the sex variable itself added signifi-candy to the predictability of salary after average matchsalaries have been used within the multiple regressionanalysis, They performed a sample survey, then matchedfaculty by department, rank, publications, degree, andacademic honors, A routine regression analysis was per-formed using independent predictors, then a stepwisemultiple regression was attempted using sex-by-predictorindices, The two interactive indices of strongest predictivepower were found to be sex-bymerit and sexby-experience. The authors were able to conclude after theirinvestigation that:

Papers read at meetings. honors, and if sex isknown, bulletins and technical reports. and yearsspent at the University of Illinois at the currentrank are significant predictors of salary (Loeb andFerber, 1971, p. 243).

The authors then attempted to place dollar amountson the degree of sexual discrimination, but found theycould only use nonstandardized regression weight ratios todefine the "average yearly dollar value of masculinity"(Loeb and Ferber, 1971, p. 246). In the authors own words,"the data (did) not dearly allow estimation of themagnitude of the discrepancies between the sexes" (Loeband Ferber. 1971, p. 243).

The Astin and Bayer investigation concentrated onthree criterion variables (academic rank, tenure status, andsalary) and four sets of predictor variables (demographic,education, professional/work activities and employinginstitution). Analytic procedures consisted of routine step-wise multiple linear regression. as cited in standard texts.The authors used a total of 33 predictors of salary andfound that "64 percent of the variance in salary (was) ex-plained on the basis of the 33 . . variables" (Astin andBayer, 1972, p. 108). Another finding deemed significant by

the authors was that:The three moss important variables in explainingsalary differentials were rank, produclivity andtypo of parent institution (Astin and Bayer, 1972,p. 108).

Significant predictors of academic rank were foundto be the doctorate, years in academe, and publications forboth men and women, In considering rank as well assalary, AstIn and Bayer added an additional demension totheir analysis with no appreciable change in analyticprocedure,

Like Loeb and Ferber, Astin and Bayer's concludingstatements could only deal with such topics as explainedvariance between the sexes and significant predictors ofpromotion and salary within each sex. These types of state-ments obviously help discern if discrimination existswithin a faculty group. but they do not allow an individual.ized study of faculty commensurate with individual facultydiscrimination allegations,

'There are other disadvantages to using the multipleregression approach besides the inappropriate indicesgenerated, The technique always requires that whenalleged bias between two groups is studied, the predictorvariables chosen be complete in their definition of pre-existing group differences (Lord, 1967), have perfect pre-dictor reliability (Linn and Wells. 1971), and be themselvesunbiased measures of accomplishment (Thorndike, 1971).In the opinion of Thorndike, all of the three qualificationsare usually substandard when used in remuneration andplacement analysis (Thorndike, 1971). Add to these quali-fications the inflation problems associated with using manypredictors to artificially raise a multiple R and the tech.nique becomes even more questionable.

In summary, there is considerable evidence to suggest that the multiple regression approach is inappropriatefor an institutional analysis of alleged sexual discrimina-tion. With reference to faculty salaries and promotionanalysis, it is not specific to members within the group, ithas a built-in bias when many predictors are used togenerate the multiple R's, the catcgores within the predictorvariables themselves are often not orthogonal, and the indices produced are not readily understood by the averageaffirmative action official,

Concerning hiring practices, the fact that many ad.iiiiiiistrators use predefined, multiple cutoff strategies inscreening applicants further prohibits the use of themultiple regression general agreement that outstandingcandidates possess areas of expertise for which there is nosubstitute. Candidates below minimum competencies inone criterion area are not generally defined as assets to acollege, even if they are superior in other areas specified aspredictors in the multiple linear regression equation,

eening Committee members who eliminate candi-

La Bay and Foster

dates lacking essential skills intuitively reject multiple Krationale, In eliminating the applications of individualswho lack one or more essential skills or experience qualifi-cation normally included as predictors of successful ac-complishments in academe, administrators reject thefoundation of regression analysis (the selection of condi.dates with the highest level of summative qualificationswho may compensate on expertise credentials with exceptional skills or experience in other cognates) (Anastasi,1968).

The best alternative approach for quantifying institutional selection and remuneration of faculty appears to be atool researchers have had for a long time, elementary Bay°sian and crosstabulation analysis, The American CollegeTesting Program's Research and Development Division,under the direction of Melvin Novick, is currently the mostknowledgeable authority on Damian techniques as appliedto selection procedures. It appears that simple, computer-assisted cross tabulation of salary and promotional re-muneration based upon accepted expertise variables is thebest technique for documentation in response to discrimi-nalion allegations. Two references recommended forreview of such method& are (1) Bias in Selection (Cote,1972), and Statistical Package for the Social Sciences(SPSS) (Nie, Bent and Hull, 1970).

Co le's research report first introduces the reader tobasic Bayesian considerations, then guides him through abrief history of selection analysis as interpreted throughBayesian theory. The analytic models considered, in orderof complexity, include the quota and regression models pre-viously discussed in this paper plus the Darlington model,The SPSS manual is an extremely powerful computer soft-ware package that is readily adaptable to institutional re-search work,

All models discussed by Cole allow the researcher touse quantifiable "priors" of his institution to predictchanges of selection for, candidates with predefined ex-pertise levels of competence. Although the bulletin iswritten specifically for admissions officers who wish toinsure equal opportunity of selection based upon achieve-ment scores, it does provide a rationale useful for insti-tutional researchers. If a university has available a campus.wide format for screening applicants, many ideas withinthe bulletin can be used directly.

If screening procedures are less formal, one may stilluse Bayesian techniques to investigate alleged discrimina-tory hiring practices. The only requirement is that thenumber of potential applicants for a position and relatedexisting conditions of current faculty of similar position areknown, For example, suppose that there is available a fileor document which defines the proportion of female tomale candidates graduating with appropriate credentials inthe field of English. Let's arbitrarily state that this propor-

153

Sexual Bias

lion is .35 to .85 for females to males, respectively.Denoting mates by M and females by F for a random

candidate, the probabilities associated with each sex maythen bo assumed to be:

P (M) e. .85

P F .35

Next, suppose that D represents an instance of a dis.crimination allegation against an institution. If that insti.tution can, through a breakdown and cross tabulation ofqualifications, salary, and promotions of current faculty,determine an index of discrimination, judgements of prog-ress toward antidiscriminatory practices can be attempted.From the index, the percent of M and F expected to bo dis.criminated against if current policy continuos can bo de.rived, and a conditional probability statement can bedefined.

Suppose, for instance, that previous statistical break.down analysis indicates a mean difference of $1500 be-tween M and F English faculty after all relevent variablesof remuneration have been controlled. Further statisticalcrosstabulation analysis Indicates that 75 percent of thefemales and 25 percent of the males were discriminatedagainst, as defined by such salary differentials. In terms ofconditional probabilities

PI DI F) .75

PI DI M .25

According to Bayes theorem, the probability that afemale candidate will be hired if interviewed (that is notdiscriminated against when credentials are identical tomale applicants) is

PIDIF)P( FI D) PID1F) PIF) P(DIM) P(M)

(.75) (.35)I (.75) (.35) + (.25) (.85)

(.282) / I (.282) + (.182))

.111

Without prior information of the extent of existingdescrimination within the English Department, it would beassumed that the probability of a female being hired is .35.With selected "priors," however, expectation of the femalecandidate is considerably reduced.

Once a university has adequately analyzed existingpromotion and salary indices, and have systematically re.viewed and defined criteria of placement and promotion,the institutional researcher should have little trouble usingeither the Bayesian or routine cross tabulation procedures.There are only two restrictions associated with these typesof analyses. The first is that all relevant criteria on allfaculty must be known. The second is that only the mostrelevant and discriminating indices (between the sexes)must bo used in initial crosstabulation work. Concerningthe latter restriction, the number of variables used inalleged discrimination cases generally should not be greaterthan four. If this number is exceeded, it is likely that thenumber of empty cells in the initial tabulation matrix willprohibit further investigations. Pragmatically, it is difficultto discuss individual competencies with more than thisnumber of qualifiers, and analytic procedures usually em-phasize the limitation of such conversions.

Anastasi, A. Psychological Testing. New York: The MacMillian Company, 1968.

Astin, H.S. and Bayer. A.E. Sex Discrimination in Academe, Educational Record, 1972, 53, 101.118.

Bottenberg, RA. and Ward, 1.H, Applied Multiple Linear Regression. Washington, D.C.: I.S. Department of Commerce, Office ofTechnical Services, AD413128. 1963,

Cole, N.S. Bias in Selection. ACT Research Report, 1972, 51.

Linn, ItL. and Warts, GE. Considerations for Studies of Test Bias. Journal of Educational Measurement, 1971, 8, 1.4,

Loeb, f. and Ferber, M. Sex as Predictive of Salary and Status on a University Faculty. Journal of Educational Measurement, 1971, 8, 1,4.

Lord, P.M. A Paradox in the Interpretation of Group Comparasons. Psychological Bulletin, 1987.68, 304.305

Newsweek. New York, N.Y.: November 6, 1972, 114.

Nie, N.H., BeM. DAL, and Hull, C.N. Statistical Package for the dal Sciences. New York, .N.Y.: McCraw-Hill Boos Company, 1970.(Note An update manual is available through National Opinion Research Center, University of Chicago.)

154

Seabury, P tU and the Universities. Commentary, 1972, 53, 28.44.

Zwerdling, 0. Sc, ( Discrimination on Campus: The Womanpower Problem. The Now Republic, 1971, 164, 11.13.

Way and Foster

155

BIAS IN FACULTY REPORTS OF TIME AND EFFORT EXPENDITURE

This paper is an extract from a detailed exploratorystudy into the bias associated with faculty self-reporting oftime and offer,. The parent study contains a substantialamount of statistical material not included here. It is thepurpose of this paper to present a general discussion of thesubject and leave the technical material to those who maywish to pursue it.' This paper, then, delves into the flues-tion of whether there is bias in faculty reports of time andeffort and whether there is a pattern to any such bias.

BackgroundIn order to put this report into perspective, let us re-

view the circumstances which led to its preparation. InJanuary, 1971, this writer was directed to determine howfaculty members at the Florida State University divide theirtime among specified categories of academic activity. Thepurpose of the project was twofold; First, it was to provideinformation for impending legislative hearings and second,it was needed to produce baseline information for the up-coming program budget and six-year plan.

A brief review of literature in the educational fieldrevealed that although there were many studies on facultytime and effort, no substantive studies had been performedon the subject of survey methods for faculty time/effort re-porting. With the absence of such research, an attempt wasmade to determine how other universities were obtaininginformation on faculty effort distribution. Letters were sentto the directors of institutional research of the forty UnitedStates universities larger than Florida State University,those forty offering the largest number of doctoral degreesin academic year 1969.70,

These letters asked for information on how each uni-versity went about obtaining information on its faculty'sdivision of effort. Of the forty schools polled, thirty replied.Of that number, 16 required a faculty questionnaire, 5 ob-tained the information from department chairmen, and 9made no attempt to obtain the information.

The results of the poll indicated that a mail question-naire was the common method of obtaining facultylime/effort information. These results led to a decision toutilize a questionnaire instrument to determine how thefaculty at the Florida State University divided its effortamong selected activities. This writer then constructed an

156

Patrick H. Sullivan, (Florida State University)

instrument that requested faculty information required bythe university. This instrument utilized the best conceptsand approaches of each questionnaire received from theother universities. The resulting questionnaire was distri-buted directly to each faculty member having a half-time orgreater appointment.

The writer then conducted faculty briefings and In-stitutional meetings campus-wide on how to complete thequestionnaire. The reaction to the questionnaire itself, thefaculty concepts of the reasons for it, their apprehensionsabout the uses or the misuses of the data, and their view ofthe process as a threat impinging upon academic freedomwere revealed as real forces at work among the faculty.These factors were seen to be significant and unavoidableforces impinging upon the reporting process.

Given these attitudes of fear and apprehension onthe part of many faculty members, this researcher began todoubt whether it is possible to obtain accurate and Un-biased information on faculty activities through use of aquestionnaire instrument. In other words, it was felt that itwould not be possible to accurately measure output of thefaculty resource.

Contributing thoughtsIn a paper presented to the American Council on

Education's 1959 Conference on Faculty Workload, Dr.Theodore Cap low recognized that there are an infinitenumber of ways to measure faculty toad, but all of thesepose problems. The first problem, he felt, is that there willbe cheating when faculty load is measured by an adminis-trative office. That this is so is not disreputable. It is amatter of general observation that whenever worker outputis measured by employers there is some misrepresentationunless extreme (and in this case, Cap low felt, inappro-priate) care is taken to verify the figures,

In commenting on the deviltries of faculty loadstudies, Reeves and Russell made the following commentthat provides an excellent summation of the situation.

The evaluation of faculty load is an extremely diffi-cult problem, Teaching duties and other professionalduties vary tremendously from institution to institu-tion and individual to individual within the giveninstitution. in fact, the factors involved in determin-

Ins total faculty load aro so numerous and so variedas almost to preclude precis() determination by anymechanical method. No thoroughly scientificmethod of measuring faculty load is now available,Existing measures are unsatisfactory and Incomplete.The answers aro not yet In Vet, as a practical neces-sity, some method of measuring and adjusting facultyload even though approximate must bo cm-ployed,3

While measurement of faculty output has been aconsistent and continual problem in higher education for anumber of years, there are no significant difficulties withmeasuring the output of processes which aro easily quanti-fied. Assembly linos producing automobile parts, electriccomponents, CIpacks of Coca Cola, or cartons of cigarettespresent no conceptual measurement challenges. Engineershave been ingenious in devising methods of measuring,weighing, and otherwise assigning numerical values to theoutputs of such operational processes.

On the other hand, wherever an output or opera-tional result involves the human resource, measurement be-comes more challenging, The Hawthorne experiments gaveevidence of this phenomenon. Under varying operatingconditions worker output Increased regardless of the changing physical environments devised by the experimenters. Itwas determined that tho presence of outsiders (the experi-menters) was itself affecting the performance of thoworkers, Studies since then have determined that either thepresence of observers or the knowledge that performance Isbeing measured may have an inpact on the amount andtypo of action taken by the person being measured.

Hypotheses

In developing tho methodology to attack tho problemof examining whether (and to what degree) there is bias inreporting time/effort through the use of questionnaires, thefollowing questions (hypotheses) were proposed:1. Is there any significant bias in reporting faculty

time/effort on questionnaires?2. Is there any significant difference in reporting bias

between academic ranks?3. Is there any significant difference In reporting bias

between academic disciplines?

MethodologyWith no existing studies to use as a baseline or point

of departure, the research techniques described herein areexploratory in nature. That is, they attempt to determine thepresence and pattern of occurrence of any bias in reportingon self-completion type questionnaires.

An ideal approach to this kind of problem of determining bias would be to compare measured responses toknown values. Since the known values did not exist, otherprocedures were needed.

Patrick H. Sullivan

As an alternate methodology to the above, tho fol.lowing wos decided to bo a satisfactory substitute, Truervalues than those available through the questionnaireswore used in place of the absolute known values. Thosetruer values wore obtained through use of another measurIng instrument more accura,a than the questionnaire,

A careful review of the literature revealed that aWary instrument, when used under carefully controlledcircumstances, will provide more accurate information ontime/effort than will a questionnaire,4

It wos therefore decided that the test instrument tobe used to provide an alternate source of information onfaculty timeleffort would be a diary log sheet. Faculty es-timates of time/effort were gathered using both questionnoire and diary instruments. In the case of the question.noire, faculty were asked to estimate their activilibs for anaverage week during the academic quarter. The diaryinstrument, on tho other hand, was administered over thecourse of the quarter such that each respondent completeda diary on each of fourteen days? selected by a computerrandom number generator. For selected activities, thealgebraic difference between questionnaire and diary esti-mates of a week's activities was obtained. This differencerepresented a measure of the bias in the questionnaireresponses.

Bias, a term used throughout this paper, is a syno.num for noise, anomaly, delay, or difference, The word"bias" carries with it no connotation of either goodness orbadness. It has no connection with statistical bias. it is thedifference between the real amount of time/effort devotedto certain activities and the apparent amount of time/effortdevoted to those activities as reported through the informa-tion channel.

Thus, the algebraic difference between a sample re-spondent's diary results and the questionnaire's reportednorm for that cell represented tho difference between theindividual's actual activity and the estimated activity forindividuals of that cell. The assumption was made that thisdifference represented a measure of bias and not the biasitself. Due to a lack of any previous developed techniquefor determining bias, this researcher felt justified in usingthe above measure. Further, it was hoped that the exploretory nature of this research provided sufficient license forexperimentation.

In view of the foregoing, a sample of faculty diarieswas obtained in such a way that it could be used for twotypes of tests: First, to determine whether the diary datawere significantly different from the questionnaire data,and second, to determine if it might be possible to detectsignificant differences in bias between academic ranks oracademic disciplines.

As for the activities to be measured, the originalquestionnaire contained a number of candidates. Selectingthe particular values to be studied from this list of condi

157

BIAS IN REPORTS

dates required the development of a set of selection criteria,A first consideration was that the value (or param

der) should be clearly defined and not easily misunderstood, A second criterion for selection of parameters wasthat they should be common to all faculty.

In order to meet the criteria of being easily under-stood and easily quantified, activities wore defined in sucha way 03 to be consistent with the faculty's perception ofthe task, Measuring activities in the category "academicsupport" would not be desirable since this represents a setof activities grouped for budgetary convenience and not aset of task performed by a faculty member, Therefore, aprofessor might not find it easy to perceive, and hence beunable to correctly quantify his "academic support" acti-vities.

It was therefore decided to measure instructionalactivities using both the questionnaire and the diary. Theactivities measured and compared were: classroom teachingactivities, individual instruction activities, and academicadvising.

Forty-eight faculty members were selected using ran-dom procedures within twelve pro-selected strata. Thesestrata or cells were arranged to include faculty from all in-structional departments. Those departments were dividedinto four academic discipline groupings:I, Professional Schools2, Social Sciences3. Physical and Quantitative Sciences4. Arts and LettersWithin each discipline grouping four faculty of each rankwere selected. An application of statistical processes to thedata gathered from these people allowed for the explorationof bias and its differences between ranks and betweendisciplines.

For each of the forty-eight faculty members partici-pating in the study. diary data wore obtained for no lessthan fourteen days during the Spring Quarter of academicyear 1970-71, For each respondent seven sample days wereobtained during the first five weeks of the quarter andseven more sample days fron the last five weeks. Each ofthe two seven-day periods or sample weeks for a re-spondent contained one Sunday. one Monday, etc., such.that each day of the week was represented.

As each diary sheet was received back from a partici-pant, it was checked and the data placed onto computercards. When all diaries were returned and the datapunched, the statistical comparisons began. The averagenumber of hours per week expended in each category as re-ported on the diary was algebraically compared with thesimilar values reported on the questionnaire, The question-flair° estimates were made for an "average" week as de-fined by the individual faculty member, The differencesbetween the two, called bias here, were subsequently oper-ated upon statistically.

158

FindingsThe first question to be answered by this research

was whether there was any statistically significant biaspresent, A simple Student's "I" lost was used to make thisdetermination, The results of this lest showed bias to bepresent in all three categories of instructional activity, In allcases the questionnaire estimates exceeded the diary tabu-lations, thus indicating an upward bias in the estimation oftime expenditure on the questionnaire,

In order to respond to the questions or hypothesesconcerning the detection of differences in bias betweenranks or disciplines, a two-way blocking design analysis ofvariance {ANOVA) model was used. In this model indi-vidual bias were considered to bo the experimental units,academic disciplines were the blocks, and academic ranksrepresented the treatments, This procedure yielded a fourblock by three treatment {or twelve cell) segmentation ofthe population.

Use of this model revealed no differences in bias be-tween tho three academic ranks of professor. assoclato pro-fessor, and assistant professor for any of the three cate-gories of activity. That Is to say, that although bias factorsexisted for each rank, mere were no statistically significantdifferences in bias between ranks.

When testing for differences between disciplines,however, there were statistically significant differencesfound in the activity category called classroom instruction,The levels of bia; were as shown in Figure 1, and the statis-tical tests were as shown in figure 2.

ConclusionsThe findings of this research lead one inexorably to

certain conclusions. The first of these is that where a ques-tionnaire instrument is used to obtain self - reported esti-mates of time and effort expenditure, there is a distinctpossibility that bias will be found in the responses.

The second conclusion is that wherever accurateestimates of timeleffort are needed through self-reportinginstruments. a diary should be used in lieu of a question-naire, The results of this study demonstrate the virtues ofthe diary where accuracy of data is a prime consideration.Diaries require respondents to write down their activitiesas they occur. No reliance need be placed upon memory orupon one's ability to estimate average time expenditure.Thus, a diary is an ex post documentation of events thathave occurred, while a questionnaire typically requests es-timates of past, present, or future events.

The third conclusion concerns the existence of biasamong groups of respondents, This study has shown thatpeople differ in their tendency to bias questionnaire re-sponses. The differences between individual biases wereexpected, whereas the differences in group bias were not.The large difference between the bias reported by Arts andLetters discipline and the other three disciplines is ample

Patrick N. Sullivan

Figure 1

COMPARISONS OF LEVELS OF BIAS

20

15 "

Bias(in hours)

10

6.1

53.8

2.1

0

ProfessionalSchools

evidence that groups of respondents can collectively, andwithout apparent collusion, bias questionnaire inputs inthe same direction and to the same degree.

A fourth conclusion may be drawn concerning thecomposition of bias. Bias, as used in this research, is thealgebraic difference between diary values and question.naire values. Given this mathematical definition, one mustcarefully question the factors that make up the bias. Theexistence of both accurate and inaccurate estimates of timeexpenditure on the questionnaire lead one to believe thatthere are two major components to bias. The first of these isthe ability, or inability. of the individual to estimate timeexpenditure. The second component, independent of one'sestimating ability, concerns the existence of a consciouseffort to alter these estimates once they have been made.

SocialSciences

ImplicationsThis study should be of interest to administrators

and managers who contemplate the use of selfcompletioninstruments to obtain time/effort expenditure data from

Physicaland

Quantitative

Artsand

Leiters

employees. Although it is clear that bias can exist in dataobtained through questionnaires, it should be equally clearthat it can occur on diaries or any other selfreportinginstrument. The nature of the diary technique is such. how.ever, that it takes a more deliberate and consistent effort onthe part of a respondent to bias a series of diaries. Thus, although one cannot expect to find bias completely absent indiaryproduced data, one can expect to find its presenceminimized.

Administrators should also be aware of the tradeoffbetween the need for accurate information and the cost ofobtaining it. A carefully controlled diary census of allfaculty would provide the most reliable estimate of facultyactivity that is possible using selfcompletion devices.

This procedure is quite costly to carry out, The costsassociated with planning for such a project: printingcosts for thousands of diary sheets, the expense of envelopes for mailing, the cost of mailing (either in postage orincreased work force for campus mail), the personnelneeded for processing completed diaries and tracking down

159

BIAS IN REPORTS

figure 2TEST OF HYPOTHESIS

1. Null Hypothesis:

There Is no significant difference that can beassignable to differences In academic disciplinesbetween the mean number of hours reported inselected categories of activity on the question.naire when compared to similar activities reportedon the diary.

2. Test Statistic:

An ANOVA model and an F-test where dfim24 anddim 36

3. Boundaries for the region of acceptance:

a. For 1)=.01b. For p=.05c. For 1)=.10

Frio= 4.38F3/36=2.86F3,36 el 2.25

4. Results of calculation of the F statistic:

S.

a. Classroom Instruction: 6.6077b. Individual Instruction: 2.1561c. Advising: 1.3419

statistically significant

Implications of the results:

a. In the activity Classroom Instruction one canreject the null hypothesis at the Peg .01, .05,and .10 levels.

b. In all other activity categories one may notreject the null hypothesis based upon thisevidence but must reserve judgment.

missing diaries, and finally, the computer costs for tabu-lation and analysis; sum up to a large dollar amount.

At some small sacrifice in accuracy but at consider-able saving in dollar expenditures, one might conduct adiary study involving a carefully selected sample of faculty,rather than a census of faculty, At even less cost and withlesser precision of results, one might consider using a ques-

160

tionnaire census (or sample) of faculty.For most applications it is neither necessary nor de-

sirable to incur the costs associated with obtaining preciseestimates of employee time/effort expenditure. If one iswilling to accept the bias likely to occur in questionnairegathered estimates of time/effort expenditure, then the pre-cision versus cost decision should be an easy one to make.

Quito simply, with decreasing' requirements for accuracy,concomitant cost reductions can be realized.

In closing I would like to share with you thethoughts of a pioneer in the field of institutional research,Hugh Stickler, The following comment from his 1959classic paper on faculty load studios provides an excellentsummary for this papor,

From the various studies to which reference hasbeen made in this paper , 'total workload cannot

Patrick If. Sullivan

simply bo described or easily measured, Literature todate is confusing, fragmentary, and inadequato. Find.ings are frequently conflicting and/or inconclusive,Relatively little real progress has been made in de.veloping adequate techniques for the measurement oftotal faculty loads' . . That was the attitude whenReevos and Russell described it in 1929; it is still thesituation thirty years later in 1959.6

And what of 1974?

'For more on this topic sea: Patrick II. Sullivan, ''An Analysis of Selected Techniques for Reporting of Time/Effort for Management andControl" (unpublished dissertation, Florida Slate University, 1972).

l'heudore Captos. "The Dynamics of Faculty Load Studies." in Faculty World Load, ed, by Kevin Bnell (Washington, D,C.: AmericrnCouncil on Education, 1990), p. 69.

IF.W. Reeves and I.D. Russell, "Instructional Loads." College Organization and Administration (Indianapolis, Ind Board of Education,Disciples of Christ, 1929), pp. 105.182

4Several papers should he devoted to this key point alone. The question of diary versus questionnaire, et. al., as it relates to this researchis addressed In Sullivan. "An analysis of Selected Techniques for Reporting ofTime/ Effort for Management and Control."

W. Hugh Stickler. -Working Material and Bibliography on Faculty Load," in Faculty Work Load, ed, by Kevin Bunnell, p. 84 citingM.E. Haggerty. The Evaluation of Higher Institutions (Chicago. Ill.: University of Chicago Press, 1937), pp. 144.58.

101

DETERMINING FACULTY ATTITUDES TOWARD COLLECTIVE BARGAINING USING PATH ANALYSIS

There has been a great deal of material publishedsince 1968 Concerning collective bargaining in highereducation, However, the majority of this literature has beenphilosophical or legalistic in nature, unsupported byempirical data. A review of the empirical research con-cerning collective bargaining in education revealed lossthan a dozen reported studios, none of which sought toanalyze faculty attitudes toward collective bargaining in afouryear college or university.

Despite the paucity of empirical research studies, thefew reported have already investigated the relationshipbetween collective bargaining attitudes and a wide varietyof demographic and career variables, While most of thosevariables were significantly related to collective bargainingattitudes, the uniform weakness of the reported re-lationships was noteworthy. One implication suggested bythese findings Is that collective bargaining attitudes arebroader in scope than mere demographic variables wouldindicate. Stronger relationships might therefore be obtainedby the identification of a variable reflective of a broader at-titudinal framework.

Levine (1985), in a study of educational viewpointsof selected suburban residents, found that a significantrelationship existed between liberal-conservatism on socialand economic issues, and views on education, In a similarstudy of the selected public school teachers by Caliguri andLevine (1967), it was found that the relationship betweenliberal-conservatism and educational viewpoint was strongenough to withstand the homogenizing effect of pro-fessional training. One implication of these findings is that:

, attitudes toward education are so embedded ina broader attitudinal framework that knowledge ofan individual's general orientation toward government and society offers some predictive powerconcerning his reaction to specific educationalpolicies.

If one assumes collective bargaining, like education,is closely associated, with governmental activity, then manyof the attitudes associated with collective bargaining inhigher education may center upon questions with broadsocial and economic implications. Therefore, facultymember's liberal-conservative orientation towardgovernment and society should be related to his collective

162

Malcolm C. McInnis, jr., University of Tennesseeand F. Craig Johnson, Florida State University

bargaining attitudes,Knowledge of the relationship between a faculty

member's liberalconservativism and his attitudes towardcollective bargaining should provide some insight into thegeneral formation of those attitudes. There is little, how-ever, that either the faculty or administration could do inresponse to this knowledge to alter faculty action, Thus,identification of a determinant over which the affectedparties have some control. such as the extent of sharedauthority seems desirable.

As early as 1967, the American Association forHigher Education (AAHE) was concerned with faculty dis-content and the potential impact of collective bargaining oncampus goverance. As part of that concern they publishedFaculty Participation in Academic Governance (1987), areport of the task force on faculty representation and aca-demic negotiations, campus governance program. That re-port, in part, found that an effective system of campusgovernance should bo built on the concept of authorityshared by the faculty and administration. Further, it foundthat shared authority can best be implemented through theestablishment of internal organizations and that formal bar-gaining relationships are most likely to develop when suchorganizations are not provided and supported.

In 1971 the AAHE published Shared Authority onCampus, another report of the campus governance pro-gram. This report, in essence, proposed a reconsideration ofauthority relationships with a view to a more effective hear-ing for students, faculty, and other inadequately heededcampus constituencies.

Other authors have also suggested the importance ofan increase in shared authority. William Boyd, President ofCentral Michigan University, suggests that unionizationresults from a ". . . search for new means of asserting pro-fessional power In areas where faculty tend to be losing tooutside forces." Similarly, Sheila Polishok (1970), of theCity University of New York, noted that;

As they probably are elsewhere, the realities of powerwere it the center of the movement toward collectivebargaining at the City University. Not unlike Ameri-can Skilled workers in the early nineteenth century,most faculty members were convinced that their pro-fessional existence could no longer bo protected by

individual initiative.if faculty discontent is related, at least in part, to a

lack of meaningful faculty participation in campus governonce, then a measure of the perception of faculty participa-tion in governance should relate to collective bargaining at.!kudos,

Previous research reporting significant, althoughweak, relationships between collective bargaining attitudeand demographic characteristic made it desirable to con-tinuo this area of investigation. Because of the low correlalions previously repelled, only the basic demographic char-acteristics of ago, sex, rank and tenure status were in-eluded, The economic emphasis of collective bargainingsuggested the need to measure faculty members' perceptionof the adequacy of the academic salary scale at their institu-tion, Recent activities of the major professional organi-zations AFT, NSP, AAUP in the area of collective bar-gaining suggested the desirability of collective informationconcerning faculty membership in those organizations.

ProcedureA research questionnaire was designed to collect in-

formation about basic faculty characteristics ago, rank,sex, tenure status, salary perception, organization member-ship together with previously developed scales to mea-sure faculty attitudes toward collective bargaining (Moore,1970), their liberal-conservativism (Levine, 1965), and per-ception of shared authority (Morgan, 1971).

The collective bargaining scale was adapted from ascale utilized by Moore (1970) in his study of PennsylvaniaCommunity College faculty. Moore's scale consisted ofthirty opinion statements concerning collective bargaining.sanctions and strikes in higher education. The only modifi-cation of his scale was the substitution of collective bar-gaining for "collective negotiation" throughout the scale.

Tho liberal-conservatism scale was originally de-veloped by Levine in a study of its correspondence to edu-cational viewpoints (1985). The scale consisted of twoparts, each with ten items, to measure liberalconservatismon social and economic issues. This scale was used withoutmodification.

The scale used for the measurement of faculty per-ception of shared authority was adapted from one de-veloped by Morgan (1971). In his exploratory study of the"Measurement of Decentralization in University Organiza-tions," the author had faculty members rate the order of im-portance of twenty-six select decisions. Subsequently,these decisions were rated by faculty using a 5-item Likert-type formal graduated to show increased sub-unit control.The only modification of the Morgan scale was the deletionof eleven of the original twenty-six decisions to more ap-propriately reflect decisions associated with the sub-unitselected for this study.

McInnis and Johnson

Analysis of the DataTho research questionnaire was distributed to all

full-time faculty members at The Florida State University,A total of 433 answer sheets, 39.30 percent, wore usable. Ananalysis of the distribution of the sample, according tobasic demographic characteristics, indicated an extremelyclose fit to the population with no significant deviations.

LiberalConservatIsmIn order to determine the linear relationship Uetween

scores on the liberal - conservatism and collective bargainingscales, a Pearson product-moment correlation coefficientwas computed. The correlation coefficient was equal to-.58, which was statistically significant at the .05 level (p<.001). Tho Power of this test for a medium effect size (ra.30) was greater than .995. Thus the inference can be madethat a linear relationship exists between faculty liberalismand positive collective bargaining attitudes.

Refinement of a linear relationship of those two vari-ables was accomplished by dividing the sample into sub-groups according to the selected faculty characteristics.Each subgroup rank, ago, sox, tenure status, salary per-ception, organizational membership was further sub-divided into classes,

Pearson productmoment correlation coefficientswere computed, for each class within the subgroups. Thocorrelation coefficient of each class together with a test ofsignificance is presented in Table 1. The results (Table 1)indicate that the correlation coefficients for all classes ex-cept two were significant at the ,001 level. The class forfaculty over 60 years of age was significant at the .05 leveland the class for AFT membership was not significant atthe level, selected (.05). Correlation coefficients substan-tially greater than for the entire sample (r= -.58) were notedfor assistant professors (ros -.62). instructors (r=.72),faculty ago 30.39 years (r= -.61), females (r= -.83) andnontenured faculty (rap -.82). Substantially lowercorrelation coefficients were associated with AFTmembership (r= -.11), AAUP membership (rut -.40), tenuredfaculty (r= -.47), professors (r= -.41) and faculty over age 60(r= -.36).

Tho comparison of correlation coefficients betweenclasses within each subgroup indicated five significant dif-ferences (.05 level). The correlation between liberal-con-servatism and collective bargaining was significantly lowerfor AFT members (r= -.11) and AAUP members (r= -.40)than faculty indicating no organizational membership, (i.e.,AFT, NSP, AAUP), (ri -.59), Non-tenured faculty (r-.82)indicated a significantly higher correlation than did thetenured faculty (r= -.47). Two additional significant differ-ences were indicated in comparisons associated withfaculty rank, Professors (r= -.41) had a significantly lowercorrelation than did either instructors (r= -.72) or assistant

163

COLLECTIVE BARGAINING

Table 1

PEARSON PRODUCT-MOMENT CORRELATION COEFFICIENTS FOR FACULTYSCORES ON LIBERAL-CONSERVATISM AND COLLECTIVE BARGAINING

SCALE BY SAMPLE SUBGROUP

Subgroup Class N r

Rank Professor 136 -.41Associate Professor 102 sp..Assistant Professor 157 -.62Instructor 27 42

Age Under 30 6131.39 Years 133 -.6140-49 Years 107 -.5250.59 Years, 95 %sp.Over 60 Years 37 -.36*

Sex Male 356 -.54Female 76 -.63

Salary Below Average 217 38.Perception Average 153

Above Average 49 -.55Tenure Not Tenured 234 -.62Status Tenured 191 -.47***

Membership AFT 32AAUP 135 -.40None 258 ..59...

Significant at the .05 level.

Significant at the .001 level.

professors (r= -.62).

Shared AuthorityIn order to determine the linear relationship between

scores on the shared authority and collective bargainingscales. a Pearson product moment correlation coefficientwas computed, The correlation coefficient was equal to-,41, which was statistically significant at the .05 level(p < .01). The power of this test for a medium cffect size(r= -.30) was greater than .995. Thus the inference can be

164

made that some relationship exists between low facultyperception of shared authority and favorable collective bar-gaining attitudes.

Refinement of the linear relationship of these twovariables was accomplished by dividing the sample intosubgroups according to the selected fa(culty characteristics.Each subgroup rank, age, sex, tenure status, salary per-ception, membership was further divided into classes.

Pearson product-moment correlation coefficientswere computed, for each class within the subgroups. The

c.

McInnis and Johnson

Table 2

PEARSON PRODUCT-MOMENT CORRELATION COEFFICIENTS FOR FACULTYSCORES ON SHARED AUTHORITY AND COLLECTIVE BARGAINING

SCALE BY SAMPLE SUBGROUPS

Subgroup Class N r

Rank Professor 136 -.31***AvIclate Professor 102 -.09Assistant Professor 157 -.05Instructor 27 -.05

Age Under 30 61 -.0531.39 Years 133 -.28"40-49 Years 107 -.27"50.59 Years 95 -.04

Over 60 Years 37 -.04

Sex Male 356 .1144Female 76 -.08

Salary Below Average 217 -.13Perception Average 153 -.12

Above Average 49 -.20

Tenure Not Tenured 234 -.02Status Tenured 191 -.30*"

Membership AFT 32 +.05AAUP 135 -.14None 25C -.08

*Significant at the .01 level.

44 Significant at the .001 level.

correlation coefficients of each class together with a test ofsignificance is presented in Table 2. The results (Table 2)indicate a significant correlation between shared authorityand collective bargaining only for professors (r0.-.31), ageranges 30.39 (v° -.a) and 40.49 (r.' -.27), males (r... -.21)and tenured faculty (r= -.30). Correlations substantially

below that for the entire sample (r= -.14) were noted for allranks except professor, females (r= -.08). nontenuredfaculty (r= -.02), AFT members ( +.05). and the non-membership class (r= -.08).

The comparison of correlation coefficients betweenclasses within a subgroup indicated only two significant

165

COLLECTIVE BARGAINING

Table 3

PEARSON PRODUCTMOMENT COORELATION COEFFICIENTS AND COEFFICIENTSOF DETERMINANTS FOR FACULTY SCORES ON THE COLLECTIVE

BARGAINING SCALE BY SUBGROUPS

FacultyCharacteristic

CorrelationCoefficient

Coefficient ofDetermination

r2

Rank -13" .02Age .20" .04Sex .14" .02Salary Perception .17" .03Tenure StatusOrganization

16" .02

Membership .35" .12

N 433

se Correlation coefficient significant at the .01 level.

differences (.05 level). The correlation between sharedauthority and collective bargaining was significantly higherfor professors (rag ..31) than assistant professors (rx. -.05).Similarly tenured faculty (rig -.30) had higher correlationcoefficients than did nontenured faculty (r= -.02)

Faculty CharacteristicsPearson product-moment correlation coefficients

were computed between the selected faculty characteristicsage, rank, sex, tenure status, salary perception, organiza-

tional membership and collective bargaining attitudes.The results including the correlation coefficeint, coefficientof determination and test of significance are summarized inTable 3. Because directionality was not hypothesized, atwo-tailed test of significance at the .05 level was selected.

Path AnalysisThe results of this "'study, as have the previous

studies, indicated a significant but weak linear relationshipbetween demographic characteristics and collective bar-gaining attitudes. Therefore, further analysis to examine themeaningfulness of these relationships was indicated.

First a hypothesized causal model was developedspecifying the interrelationship between the independentvariables rank, age, sex, salary perception (SP), tenurestatus (TS), liberalconaervatism (LC), shared authority

166

(SA), organizational membership (OM) and collectivebargaining attitudes. Second, the causal model was testedusing the techniques of path analysis.

The results of that analysis according to the hypothe-sized causal ordering are summarized in Figure 1. Thedirect effects between determining and the dependent vari-ables are shown by Weigh line arrows. The value ac-companying those arrows is the path coefficient (i.e., stan-dardized regression coefficient). Tho curved arrow betweenage and sox indicates an unanalyzed correlation. Thearrows coming from outer space denote the residual (i.e.,the variance unexplained by the multiple regression equa-tion).

According to the hypothesized causal model, certainrelationships are assumed. First, sex and age influencedrank. Both variables were found to have significant path cot/efficients with age being the stronger of the two.

Second; rank, age, and sex influenced tenure status(TS). Only rank and age had significant path coefficientswith (TS). Substantially, this finding says that sex does nothave a direct bearing upon (TS) but does have an indirecteffect through its influence on rank. With sex removedfrom the regression equation new path coefficients werecomputed indicating rank somewhat stronger than age.

Third; rank, age, sex and tenure status (TS) in-fluence salary perception (SP) of these variables only sex

Figure 1

PATH ANALYSIS SUMMARY

.11148

SE X

AGE

RANK

.947,

.137 LIBERALISMCONSERVATISM

(LC)

tP)

.12S

.238

SALARYPERCEPTION

(SP) ,

SHAREDAUTHORITY

(SA)

McInnis and Johnson

.9450

ORGAN.MEMBERSHIP

(OM)

.301

.767S

.177

.113

.198

371

TENURESTATUS

(TS)

.144.172

223

097

.7911

was found to have an insignificant path coefficient. Thismeans that sex does not have a direct effect on (SP) otherthan its transmitted through rank directly to (SP) or Ind]rectly through (TS). The new path coefficients computedwith sex deleted indicated that rank had a stronger influence than either age or (TS).

Fourth; age, rank, sex, tenure status (TS), and salaryperception (SP) influence liberal-conservatism (LC). Onlyage and sex were found to have a significant path coeffi-dent with (LC). This means that rank, (TS), and (SP) donot have a direct effect on (LC). The new path coefficientswith rank, (TS), (SP) deleted indicated that age has agreater influence.

Fifth; rank, age, sex, tenure status (TS), and salaryperception (SP) influenced shared authority (SA). Theregression analysis indicated that age, rank, and (SP)directly influenced (SA). This means that the influence ofsex is indirect and only to the extent it affects rank whichaffects (SP).

Sixth; rank, age, sex, tenure status (TS), salary per-ception (SP), liberal conservatism (LC), and shared

.9662 .9810

C0L

L

C

VE

AR

A1

N

ATT

1

U

S

authority (SA) influenced organizational membership. Theregression analysis showed that of these variables only(LC) and (SA) were directly related. The remaining vari-ables had indirect effects through these two variables. Thevarious paths were discussed under the section dealingwith (LC) and (SA) as dependent variables.

Lastly, all of the independent variables, rank, age,sex, tenure status (TS), salary perception (SP), liberalcon-servatism (LC), shared authority (SA), and organizationalmembership (OM) influenced collective bargaining atti-tudes. The regression analysis showed that age, (LC), (SA),(SP), and (OM) had significant path coefficients with col-lective bargaining. This means that sex, rank and (TS) in-fluence collective bargaining attitudes only to the extent oftheir indirect influence on (LC), (SA), and (OM). When thepath coefficients were recomputed with rank, sex and (TS)deleted, the significant variables, in order of the strength oftheir effects were (LC), (SP), (OM), (SA) and age.

The results of path analysis indicated that sex, rank,and tenure status do not have a direct effect on collectivebargaining attitudes. According to path theorem, if a truly

167

DETERMINING FACULTY

nonsignificant path 13 deleted, Iho recalculated correlationwill equal Iho original zero order correlation. To test thedeleted paths, rank, sox, and tenure status, the correlationcoefficients for the entire matrix wore recalculated. Table 4presents the correlation matrix with the original coefficienton top and he recalculated coefficients on the bottom. Acceptablo differences (loss than .05) between the originaland recalculated correlation coefficients wore indicated forsex, rank, and tenure status, A comparison of the differencebetween original and recalculated correlation coefficientsfor tho entire matrix suggests a reasonably close fit for thehypothesized model.

SummaryAnalysis of the liberalconservatism scale revealed a

significant positive correlation between faculty liberalismand favorable collective bargaining n ides. The strengthof this correlation was weakest let faculty who were over60 years of ago, tenured, of the rink of profes. a

member of AFT or AAUP, When rrelations orecompared between groups, significant difference, ,,renoted for the following groups: professorsinstrui

168

professorsassistant professors, tenurednot tenured,AAUPnone, and AFTnone.

Tho analysis concerning the shared authority scalerevealed a significant correlation botweon a low perceptionof department autonomy and favorable collective bargain.Ins attitudes. Tho strength of this correlation was increasedfor faculty who wore a full professor, male, or tenured,When correlations wore compared botweon groups, asignificant difference was noted botweon professor.assistant professor and tenured non tenured groups.

The results of the path analysis indicated that sox,rank, and tenure status do not have a significant directeffect on collective bargaining attitudes. Tho effect of agoon collective bargantng attitudes was found to bo primarilyindirect, although a small direct effect was noted.

ConclusionTho findings suggest that factors contributing to

faculty attitudes toward collective bargaining may bo morecomplex than has been generally acknowledged. Thoreliance upon basic faculty demographic characteristicsmay bo limited to describing the proponents of collective

TABLE 4

CORRELATION MATRIX INDEPENDENT VARIABLES WITHCOLLECTIVE BARGAINING FOR ORIGINAL ZERO

ORDER CORRELATION COEFFICIENTS (UPPERPART) AND RECALCULATED CORRELATION

COEFFICIENTS (LOWER PART)

IndependentVariables Sea Age Flank tenure Salary

SharedAuthority

liberalConservatism

OrganizationalMembership

Collective ear-gaining Alta.

Sea 1341 .1293 .0529 .14681 .0443 -.1779 .0801 .1409

Age 1541 -3384 .S2S7 -.1701 .1111 .0090

Man', .F199 .5384 -.3464 .0297 .03S8 .1190 .02SS 1288

Tenure 0023 .52S5 3464 -.1311 .0414 -.1114 .0404 .1577

Salary 0736' -.1040 .0364 .3341 0191 .0711 .0688 .1728

SharedAuthority 04% .1110 .0284 .0347 -.1291 .0286 -.t473 -.1391

LiberalConservatism -.1778 -.nes .1264 1479' .0556 -.1874

OrganizationalMembership 0442 .0653* .0431 .0357 .0039' -.1641 -.2963 .3525

Collective&Alga TilingAttitudes .1021 .2053 -.0829 .1728 3708 .1677 -.5661 3453

Difference between original zero order correlation and recalculated correlation greaterthan .05.

bargaining rather than explaining their preference. Ago,sex, rank, and tenure status appear to affect collective bar.gaining attitudes primarily in an indirect way and thushave limited use as predictors.

'rho findings concerning the liberal conservatismscale suggest that faculty attitudes toward collective bar.gaining are broadly based and generally similar to anindividual's view of government and society, Refinedanalysis of this variable suggests that specific situationalvariables tend to override this general relationship.

Tho relative weakness of the relationship of theshared authority scale to collective bargaining attitudes

McInnis and Johnson

suggests that emphosis on this determinant may bomisplaced. it should bo noted, however, that this variable isrelated to departmental autonomy, 'rho findings suggestthat the strength of this relationship may bo depressedbecause of differences of individual autonomy within de-partments,

The results concerning salary perception generallysupport the emphasis of collective bargaining on economicissues. However, the weakness of this relationship suggeststhat reliance on this determinant as being primary may notbe justified.

American Association for Higher Ecucation, Faculty Participation in Academic Governance. Washington D.C.: American Associationfor Higher Education, 1967.

Boyd, William B. "Collective Bargaining in Academe: Causes and Consequences," Liberal Education Vol. 57 (October, 1971).

Keeton. Morris et al.. Shared Authority on Campus. Washington, D.C.: American Association for Higher Education, 1971.

Levine, Daniel U. "Socio-Economic Attitudinal Corelales of Educational View Point," The journal of Experimental Education, Vol. 33(Spring, 1965)

Levine, Daniel U., and Catigura, Joseph. "Relationships Between Teachers' Views on Education and Their SocioEconomic Attitudes,"The Journal of Experimental Education, Vol, 35 (Summer, 1967).

Maid, Michael S., And Jacoby, Jacob. "Is There an Optimal Number of Alternatives for Liker Scale Items? Study I Reliability andValidity," Educational and Psychological Measurement, Vol. 31 (Autumn, 1971).

Moore, John W. "The Attitudes of Pennsylvania Community College Faculty Toward Collective Negotiations In Relation to Their Senseof Power and Sense of Mobility." Unpublished Ph.D. dissertation, The Pennsylvania Slate University, 1970.

Morgan, William E. "Measurement of Decentralization in University Organizations," American Educational Research Puma!, Vol. 8(March. 1971)

Poinhok, Sheila S. "Collective Bargaining and the City University of New York" loam°, of Higher Education, Vol. 41 (May, 1970).

A,

169

AN ANALYSIS OF THE EFFECTS OF TUITION AND FINANCIAL AID POLICIES

This paper describes an analysis of the offect(s) andfeasibility of changes of tuition and financial aid policieswhich may bo implemented to guarantee opportunity for acollege education to all qualified college-age youth. Thoeffecl(s) of such policy changes are described in terms ofexpected changes in college enrollments, classified byincome distributions of college enrollments, which resultfrom changes in the "unmet costs" borne by students fromdifferent family income groups. Such "unmet costs" are de-fined as the sum of the expected parental and studentcontributions towards college expenses and financial aids(excluding loans and employment).

introductionIn its regular meeting of December 1970, the State of

Illinois Board of Higher Education approved an increase oftuition charges for public senior institutions. This increasewas the first of a series of significant tuition increases to boexperienced in the early seventies.

These drastic increases in tuitiols have generatednumerous discussions on the role of public education andthe effect of high costs on the ability of all qualified youthto obtain a college education. "Equal opportunity" has be-come the central issue in all discussions of tuition levelsand financial aid programs. Tho concept of "equal oppor-tunity" has been described by Hansen and Weisbrod in thefollowing manner:

First and foremost, ono of society's major goals isthat of promoting greater equality of opportunity, agoal that can be achieved by making post-secondaryeducation and training more generally available. Thismeans that all students wishing to should bo able toattend a post-secondary school for which they arequalified, and to do so without undue financialconcern.t

"Equal opportunity" has been an integral part ofAmerican democracy, which has led to the large growth ofpublicly-supported institutions of higher education in theUnited States, especially since 1945. II is this concept whichhas resulted in low tuition levels, although evidence doesnot shown that "equal opportunity' has been achieved.

The Tuition ControversyThe trend towards higher tuition levels at public in-

170

Robert Winter, University of Illinois

stitutions has renewed the debate as to whether low tuitionlevels should bo maintained or whether, in fact, higher tui-tion levels may facilitate the achievement of "equal oppor-tunity" by allocating a part of the additional Income tofinancial aid programs which would remove financialbarriers from college education.

Three significant linos of thought have been c'e-veloped to support high tuition levels:A. Tho subsidy argument Proponents of higher tuition

levels are well represented by the following statementmade by Bolton:Some argue strongly that the present systemexcessively subsidizes higher income families,because the low tuition in public institutionsis available to students without considerationof whether they could pay their own way.2

It follows then, that by increasing tuition, funds wouldbo made available to allocate to financial aid programswhich would make "equal opportunity" objectivespossible. Tho net effect would bo that of shirting coststo higher and higher middle class families whilelowering the costs of a college education to low andlower-middle class families.

B. The viability of private education argument Low tui-tions have created a concern over their effects on theviability of private institutions. This concern wasexpressed in the "Higher Education in the UnitedStates: The Economic Problem" Seminar held in 1958:

The private colleges fear that tax-supported,low tuition state universities may eventuallyrun them into bankruptcy.

Higher tuitions at public institutions would insure theviability of private institutions since the later wouldbecome price-competitive.

C. The benefits argument Another view in the tuitiondebate has been presented by those who havediscussed college education in terms of societal andpersonal beneifts. Owen exemplifies this argument inthe following statement:

Many economists, and others. believe that thesubsidy of college students is justified on thegrounds that without subsidy, the privatevalue of a youth's college would typically beless than its social value thus leading to a

less than socially optimal level of expendilutes on college education!

Owen also asserts that both private and social benefits andcosts will be different for different groups., for example, fornonwhites the private value may bo lower and the sodalbenefit may be higher, This argument supports policieswhich increase subsidies to low and lowermiddle classyouth.

Proponents of low tuition levels argue that high tui-tion levels would destroy the illusion that education can behad cheaply, therefore, discouraging low and lowermiddleclass youth from obtaining a college education,

In the final analysis, a decision of low versus hightuition is based on the political factors affecting the stateand, more importantly, the resources available to the state,The most important factor will be the willingness of societyto support programs which will assume the achievement of"equal opportunity" objectives.

Analysis of Tuition Financial Aid ModelsThe issues discussed In the previous section raise

several concerns as to the mechanics required to achieve"equal opportunity."Whether tuitions should remain low orshould be increased, or whether societal and personalbenefits should be accounted for by the pricing structure ofcollege institutions are not treated hero as the principalissues.Tho achievement ()rogue! opportunity" is treated asan economic Issue, with attention to: plans that recognizelimited public resources, the likelihood that tuition levelswill be increased and larger subsidies may bo given to stu-dents from low and lower middle income families, and thatthese steps will be taken not on the basis of the social catenon of who deserves assistance, but by the economic edterion of who needs it most.

The effect of tuition and financial aid policies isevaluated in terms of the impact of college-ago youth andtheir families. The decision to attend college may be affected by the following factors, as described by Hansen andWeisbrod:

The cost of a college education to a student and hisfamily apart from the income foregone can beanalyzed in two parts. One can be termed the "price"of education tho tuition charge, the books and sup-plies, and so forth. The second is the "ease of finan-cing" that price that is. the availability and termsof loan funds and scholarships.3

The financial aid available to needy students may beof three form. gift aid, part-time employment, and loans.Gift aid may be defined as a nonrepayable subsidyawarded directly tc the student, while a loan will requirerepayment of the original amount plus added interest. Theanalysis presented in this paper assumes that only gift aidwill be considered as an effective tool to help achieve"equal opportunity," while loans may be used by students

Robert Winter

and/or parents to finance the expected contributions to thecost of education and the unmet cost the portion ofcollege costs not covered by gift aid,

During 1984, the Illinois Board of Higher Educationconducted a study to analyze the composition of thostudent body in institutions of higher education in the Stateof Illinois, This study concluded that:

Based on high school rank in graduating classes,25% of those students in the upper quarter and 44%of those in the second quarter do not go on to col-lege.°

The results of this study gave impetus to the rapidgrowth of the Illinois State Sch warship Commission(ISSC). This growth is consistent with the direction recommended by the Board of Higher Education in its attempt tcachieve "equal opportunity" in the State of Illinois. The ob*lives of the Illinois State Scholarship Commission areexpressed in its 1970 Anneal Report:

. , . to remove financial barriers to college attend-anus; to expand college choice; and to permit thou-sands of students to attend non-public colleges whowould have otherwise attended public colleges andadded to the state subsidy of operating and capital ex-penditures.?

The Illinois State Scholarship Commission (ISSC)offers two major financial aid programa:1. Non repayable grants awarded to "needy" undergrad

wit° students, and2. Guaranteed loans granted to students through commer-

cial lenders.The scholarship and grant programs are limited to full-timeundergraduate students who are residents of the State ofIllinois and are enrolled in-state in an institution of highereducation. The Higher Education Student Assistance Lawpermits the ISSC to grant awards up to either $1,200 peryear or tuition and fees, whichever is less.

An analysis of the distribution of ISSC awards raisesquestions over the limitations of the Higher Education Stu-dent Assistance Law. Figure 1 shows the 1972.73 distri-bution of unmet costs for ISSC recipients attending publicand private senior institutions. This distribution brings toquestion whether the ISSC objectives, stated previously,have been achieved,

In order to analyze the effect of improvements of theISSC programs, a computer simulation program was de-veloped to evaluate changes in the following policies:

1. The tuition. level for the public senior institution wasincreased up to three times the 1970.71 !eve's,

2, The extent of coverage of ISSC awards was variedbetween the current tuition and fees limitations up to acoverage of 100% need beyond tuition and fees, and

3. The maximum size of ISSC awards was varied betweenthe current $1,200 and $3,000,

The computer simulation program subdivides the

171

TUITION AND FINANCIAL AID

1,800

1,400

1,200

1,000

800

600

400

200

Figure 1

UNMET COST DISTRIBUTION

Private Senior

Public Senior

0 2 4 8 8 10 12 14 18

Parental Income(Dollars x 103)

college population into twelve parental income categories,and estimates for each category the amounts of financialaid and unmet cost for a given set of tuition and financialaid policies. An "equity index" was also calculated andwas defined as the standard deviation of unmet costs. Thisindex provides a measure of the achievement of "equalopportunity."

The development of the model was divided intothree parts:A, Description of the College Population of the State of

Illinois

172

1 1 1

18 20 22 24 28

The college population was described in terms ofincome distribution, choice of institution, and expectedparental contributions toward college expenses, Thisinformation was not available at the time, resulting inthe following approaches:1. Income distribution The most reliable informa

tion was found from the freshmen surveysperformed by the American Council of Education(ACE).

2, Expected parental contribution the expectedparental contribution towards college expenses

was based on the College Scholarship Service 1071estimates, The following assumptions were maderelative lo the number of dependents by familyincome;a. Families whose incomes are less than $10.000

average three dependents, andb. Families with incomes above $10,000 average

Iwo dependents,

11, Writing of the Computer Simulation ProgramThe program used as inputs data which described;

public senior, public junior, private senior. and privateJunior institutions (i.e., tuitions, college budgets, andfull time enrollments) and the college population ateach type of institution (1.0., the information providedby ACE and overage gift aid other than ISSC). The out-puts of the computer simulation program are summar-ized In Table 1.

C. incorporation of Price Elasticities of the Demand forHigher Education

Price elasticities of the demand for higher educationhave been estimated in works by Miller,e Campbelland Siege1,9 Feldman and Hoenack,inlioenack,32 and Zorazzini.,, Miller, for example.arrived at the conclusion that low income groups had arelatively elastic demand for higher education withrespect to price, while high income groups had an in.elastic higher education demand with respect to price.Although not necessarily representative of the behaviorof college youth and their families in the Slate ofIllinois, the results obtained by Corazzinl, at a!, wereutilized in this analysis.

Corazzini, et al. estimated price elasticities by[ining separate regressions on samples stratified bythe Project TALENT by SES variable, They arrived atthe following estimates:a. Ea= -.32 for the low SES andb. 1-'4, tit .,18 for the high SES.

The incorporation of price elasticities of demandfor higher education into the mooeis previouslydescribed in this paper permits the evaluation of theeffect of changes of tuition and financial aid policieson the total enrollment of the state and on theenrollments by institution type. Changes inenrollments by income group and by institution typewill be based on changes of the "price" of education.

The "price" of education for an individual was de-fined as the sum of the expected parental and studentcontributions towards college expenses, and the unmet

wascost. Since, as was shown previously, unmet costs willvary for different income groups as tuitions for publicsenior Institutions and financial aid policies are varied,two effects on enrollment can be expected:

1, An increase in the price of education This increase will result from an increase in tuition

Robort Winter

which must be absorbed by the student and hisfamily. This condition may he mcre typical forupper middle and high income families. Ttioexpected effects on enroltmeni are explained byHopkins' definitions;a. "Substitution' effect some students (09%

of attrition), From public senior institutionscould shift to private Institutions, and

b, "Not discouragement" effect Somestudents (31% of :ifirilion), from public seniorInstitutions would drop out of collegealtogether.

2. A decrease in the price of education Anincrease of financial aid, in the form of monetarygrants anti tuition and lee waivers, to low and middleIncome college youth will decrease thole price ofeducation. This will result In an increase in theirenrollments in institutions of higher education.

Table .2 describes the effect on students alltedIngpublic senior institutions In the Stale of Illinois ofincreases in tuition levels coupled wit': improvementsin financial aid programs offered by the ISSC. Thefollowing set of policlas Is investigated:1. Tuition level at public senior institutions: two and

half times 197071 levels.2. Maximum awai d $3.000.3. Percent coverage beyond tuition and fees 100%.

Two distinct ,ffecta can he identified:1. Unmet costs are decreased fow low and middle

income families. The new map of unmet costsresults from a significant improvement of theIllinois State Scholarship Commission financialaid polii !es.

2. The expected income distribution of enrollmentsin public senior institutions changes in twosignificant ways:a. Enrollment of youth from low and middle

income families Increases (4,554 new studentsfrom families earning below $10.000), as aresult of improved financial aid policieswhich lower the price of education, Thepercentage of enrollments from familiesearning below $6,000 increases from thecurrent 7.2% lo 8.6% of the total enrollment inpublic senior Institutions.

b. Enrollment of youth from upper middle andupper Income families decreases (3.316students from families earning above $10.000),as a result of Increases in tuitions which raiseIhe price of education. The percentage ofenrollments from families earning above$25,000 decreases from the current 9.0% to8.6% of the total enrollment in public seniorinstitutions.

173

TUITION AND FINANCIAL AID

Table ISUMMARY of OUTPUTS OF THE COMPUTER SIMULATION PROGRAM

116 MN

TUITION a To (1970-71) TUITION , 24 To TUITION so 23 To TUITION . 3.0 To

M C TR N M C Ti N M C TR N M C TR N

$ 40 $2700 100% $40.0 6-71.7 $2700 100% $11001 $-44./ $2700 100% $100.0 $44.4 $2700 100% $120.0 $-22.5

$100 $1400 100% $40.0 $ 65.6 $1800 100% $00.0 $ ns $2400 80% $100.0 $ 27,2 $2400 80% $1200 $ 16,1

$150 $1800 80% $40.0 $ 59.1 $1800 80% $00,0 $ 33,1 $1100 80% $100,0 $ 22.2 $1100 100% $120.0 $ 13,6

$200 $2100 60% $40.0 $ 51.2 $2100 60% $00.0 $ 254 $2100 60% $100.0 $ 14.7 $2100 60% $120.0 $ 3.8

$250 $1500 60% $40.0 $ 43.4 $1500 60% $00.0 $ 17.1 $1500 60% $100.0 $ 6,9 $1500 80% $120.0 $ .8

$300 $1800 40% $40.0 $ 36.9 $1100 40% 00,0 $ 11.1 $1000 40% $100.0 $ 1.1 $1100 40% $120.0 1+9.6

$360 $1200 40% 540.0 $ 27.5 $1200 40% 00.0 $ 2.4 $2100 20% $100.0 1 + oa $2100 20% $120.0 $ 18.5

$410 $1503 20% $40/1 $ 21.3 $1500 20% $00.0 1 + 3.5 $1500 20% $100.0 $ 13.9 $1200 40% $120.0 $ 23.0

$470 $1800 0% 140.0 $ 14.5 $1000 0% 00.0 $ 9.9 $1800 0% $100.0 $ 20.1 $1800 0% $120.6 $ 30.4

$520 $1200 0% $40.0 $ 5.6 $1200 0% 00.0 $ 18.8 $1200 0% $100.0 $ 29.0 $1200 0% $120.0 $ 39,3

E Equity Index; the standard deviation of unmet costs.

M Maximum award value.

C Extent of coverage the percentage of financial need minus tuition andlees by which awards may exceed tuition and fees.

TR Tuition revenues from public senior Institutions. ,

N Net profit or loss to the State; the difference between tuition revenuesfrom public senior Institutions and ISSC expenditures.

Table 2

Combined Effects of Tuition and Financial Aid Policieson Total Price and Enrollments

type of Institution: rblic Senior

---.FAMILY INCOME

UNMET COST 1 TOTAL PRICE' 1

PLAN 1

% CHANGE IIN MICE

FALL 1970 IENROLLMENTS

CHANGE IN 1

ENROLLMENTSNEW 1

ENROLLMENTS IPRICE

ELASTICITY,

PLAN A PLAN II PLAN A

0000-39994000.59996000-79990000-9999

10000-1249912500.1499915000-1999920000-2499925000-29999300003499935000-39999over 40,000

$11081110

1183

52044

0000000

$ 42S2

11754440000000

$160116101638165019392100210021002100210021002100

$ 542SS2

072118420922550255025502550255025502550

-4646.47-.28

+ Ai+.21+11+11+.21+11+.21+ .21

3,3015,995

10,42517,59325,02122,28421,89311,7294,5612,9971,4432,737

+ 71S+ 1,117+ 1,371+ 1,281

4801,030

920- 443

72

113SS

103

4,1037,102

11,7%18,87424,54121,25420,97311,286

4,3192,8841,388.r.634

-.32-.30-.28.16-.24-.22 ,

-.20-A 8.10-.18.111

10

I 4

130,0661 ,

+ 1,238 131,304

S Total Price Unmet Cost + Expected Parental and Student Contributions

PLAN A (current)

Tuition Leach T. (1970) 2.3 T.Max $1200 3000Financial Aid Beyond Tuition and Fees 0% 100%

PLAN 1

179

TI-.e results of this analysis point towards financialaid programs aimed at achieving two essentialobjectives:1 Equalizing the burden for students currently

attending institutions of higher education, and2, Reducing or eliminating financial barriers to

academically able college-age youth, who, other.wise, would not consider enrolling in college,

However, those results neither support nor relect low orhigh levels of tuition. The objectives stated above could beachieved just as well by maintaining low tuition levels andincreasing the state investment in the programs of the ISSC,

A Challenge to Current Policy Trende

In the past few years, largo Impetus has been givento those attempting to resolve the financial problems facedby institutions of higher education and collego.age youth, toimplement policies which combine increase of tuitionlevels and the expansion of financial programs. Certainly,such policies appear to be feasible, as is shown in the pre-vious section, but, in the opinion of the author, they aresubject to challenge.

Tuitions at public institutions have been treated er-roneously as the price to purchase an education, althoughtuitions have been related to the cost of education on an ar-bitrary basis only. Tuitions at public institutions aro, infact, a form of user taxes and public institutions of highereducation act as agents for the state as tax collectors, If tuitions aro treated as taxes, then it follows that their levelsshould not be established on an arbitrary or philosophicalbasis, but should be tied to the state's tax stra--iure whichrelates financial requirements to the commitments or needsof public servia.

In the same vein, financial aid programs must betreated as a service provided by the state, The decisions onthe level of support of this service should be based on poli-tical interest groups. All play a rote in making decisionsabout the shape and position of unmetcost maps. The flowof funds, such as tuitions or user taxes, from studentsthrough public institutions of the stato government, the

Rob(Irt Winter

support to institutions of higher education from the state tothe institutions and indirectly to staents, and the supportof financial aid programs from the state through an appro.priato agency to students is described in Figure 2,

The proposed definition of luitien as taxes reducesthe value of the subsidy and the viability of private educelion arguments since the elate government has the option tolegislate the level of subsidy given to each citizen for eachtype of public service, The level of support to financial aidprograms may be determined by using the benefits argu-ment, or the bask premise that all qualified citizens aro en.titled to a college education with the absence of financialbarriers. This approach would allow the achievement of"equal opportunity" oblecttves independent of tuitionpolicies.

ConclusionDuring the last few years, rising costs of education

and limited government resources have resulted in pros-sures to increase tuitions at public insitutions. Impetus toIncrease tuitions at public institutions has come from thosewho believe that low tuitions are not sufficient to insure an"equal opportunity" to obtai , a college education; twoproblems aro identified:1. Students from high and highermiddlo class families

receive a largo subsidy which they do not need, and2, Private institutions risk financial bankruptcy since they

aro unable to compote on a tuition basis.Proponents of higher tuitions recommend that in-

creases in tuitions should be coupled with increases infinancial aid programs. The expected result would be togenerate additional funds for the operation of institutionsof higher education and funds required to obtain "equalopportunity" objectives through expended financial aidprograms.

Although the oposals are feasible from an implementation view, the e not realistic from an economicand political view, Policies on tuition levels and financialaid programs should be independent of each other andshould not be combined.

175

TUITION AND FINANCIAL AID

Figure 2

FLOW OF TUITION INCOME AND FUNDS TOSUPPORT FINANCIAL AID PROGRAMS

STATE GOVERNMENT

TAX REVENUES

PERSONAL USER

INCOME SALES TUITIONS

STATESUPPORT

TOPUBLIC

INSTITUTIONS

_____YPUBLIC

INSTITUTIONS

A

TUITIONS(USER

TAXES)

TUITIONINCOME(TAXES)X ES)

V

STATESUPPORT

TOPRIVATE

INSTITUTIONS

PRIVATEINSTITUTIONS

TUITIONS(PRICE)

USERS OF HIGHER EDUCATION

176

PUBLICSERVICES

AIDPROGRAMS

STATESUPPORT

TOSTUDENTS

Robert Winter

'Hansen, Lee and Weisbrod. Burton A., "A New Approach to higher Education Finance." in Financing 'fisher Education: Alternutivesfor the Federal Government. M.D. Orwlg, Editor, The American College 'resting Program, Iowa City. 1971. p, 119.

ABollon, Riser E,, The Economics and Public Financing of Higher Education: An Overview," In The tconomics and Financing ofHigher Education In Iho United States, Wang Ion, D,G, 1969, p. 66,

iSeminar "Higher Education in the United Slates: The Economic Problem," Review of Economics and Statistics (August 1900). 42, No,3, p. 37.

40wen, John D., Towards a More Consistent, Socially Relevant College Scholarships Policy, The John Hopkins University, Baltimore,Maryland. fanuary 1970. p. 2.

oliansen, Lee and Weisbrud, Burton, A., -Tho Search for Equity In Provision and Finance of Higher Education," in The Economics andMc/acing of Higher Education in the United States, Washington, D.C., 1969. p, 109.

bit Muster Plan for Higher Education. The Illinois Board of Higher Education, fitly 1964, p. 22.

'Report. Illinois State Scholarship Commission, May 1970, p. 1.

*Miller, Leonard S., Demand for Higher Education in the United States, State University of New York at Stony Brook, New York, 1971,

oCampbell, Robert and Siegel, KN., ''Demand of Higher Education in the United Slates," American Economic Review (lune 1967), Vol.57. pp. 482.489.

loFeldman, Paul and Hoenack, Stephen A., "Private Demand for Higher Education In the United States," In The Economics andFinancing of Higher Education in the United States, Washington. D.C., U.S, Government Printing Office, 1969, p. 375.398,

"Hopkins. Thomas D., The Provision of Higher Education: A Market Interpretation, Yale University, New Haven, Connecticut. 1971,

litIoenack.Stephen A., Private Demund for Higher Education in California, Office of Analytical Studies, University of California. 1967.

14Corazzini, Arthur, Dryan, Dennis J., and Grabowski, Henry, "Determinants and Distributional Aspects of Enrollment in U.S. HigherEducation," Journal al lumen Resources (Winter 1072). Vol. 7, No. 1, pp. 39-50.

177

THE EDUCATIONAL ATTITUDES AND ORIENTATIONS OF UNDERGkADUATES:A NEW MEANS OF ASSESSMENT

Tho research in ported in this paper is focused on theattitudes and orientation which students havo toward var-ious philosophies, purposes. and processes of a collegeeducation, Even to casual observers of the higher educationscone, it is well known that our colleges and universitiesenroll individuals with tremendously different back-grounds, Interests, and abilities. As Cross (1971) indicates,institutions are faced with an increasingly heterogeneousstudent population. What are the interests and learningorientations of our students? What are their attitudes aboutthe curricular-instructional policies which affect them?What kind of educational processes do they desire? More-over, what relationship exists between students' educe-tional attitudes and their patterns of growth its both the cog-nitive and non-cognitivA domains? These havo been andwill continue to be major questions at institutions of higherlearning, although to date little effort has been focused onthese particular areas of inquiry.

This is not to say that little has been done in the areaof research on college students. To the contrary, Feldmanand Newcomb (1969) have reviewed and compiled a mas-sive number of research studies which have been under-taken over the last four decades. Mechanisms to existfor determining general characteristics and attitudes ofstudents the American Council on Education's Survey ofEntering Freshmen and the College Student Questionnaire(Peterson, 1985) are prime examples. Research on studentperceptions of the college environment performed by Pace(1903) and Stern (1983), personality characteristic studies(Heist and Yonge, 1968), student "satisfaction" studies byPervin (1967), Betz, et. al. (1971) are valuable in theirown right. However, these types of research studies do notnecessarily involve consideration of the implications of cur-ricular-instructional situations since for the most part, theresearch inventories employed do not deal directly withstudents' attitudes regarding educational processes andpolicies.

Thus. the first task in this project was one of devis-ing an instrument appropriate for our purposes, namely,providing a way to assess students' attitudes regarding theirpreferred modes of learning, their views on student-facultyroles in educational decision-making, and so on, After ex-tensive pre-testing of items and factor analyses (principal

178

Barry R. Morstaln, University of Deaware

components). these efforts culminated in the developmentof the Student Orientations Survoy (Gray and Morstain,1970) and this inventory was used in this inter-institutionalresearch study. There are ten scales in the Student Oriente-lion Survoy (SOS) with each scale consisting of eight items.Each item has four Likert type response categories, rangingfrom "not at all like my attitude" to "closely reflects myaltitude."

Before presenting some attitudinal profiles of stu-dents in different institutions and/or different degree pro-grams. it would be helpful to give a brief overview and des-cription of the SOS scales., The SOS assesses five majordimensions or areas of student orientations to college, andas noted below, each dimension is comprised of two cor-responding scales. Tho dimensions and scales are asfollows:

Table I

STUDENT ORIENTATIONS TO COLLEGE

" Preparatory" DIMENSIONS "Exploratory"Orientations (Areas of Orientations

(S Scales) Orientation) (S Scales)

Achievement 1. PURPOSE inquiryAssignment 2. PROCESS IndependentLearning Study

Assessment 3. POWER Interaction

Affiliation 4. PEERS Informal Assoc.

Affirmation S. PUBLIC In. ilvementPOSITION

Given the pattern of scale intercorrelations, one setin five scales clustered together in what has been inter-preted as a general "Preparatory" orientation to college, andanother set of five scales were related to a general "Ex-ploratory" orientation to college. That is, it appears thatwhile college is most highly valued by some for itspreparatory function in terms of acquiring useful know-ledge, skills, vocations, and social rotes it is valued mosthighly by others for its expiatory possibilities la, forthe opportunities it affords for exploring one's interests,ideas, and personal identity. These general orientations be-come more apparent when one examines the content and

Sid.Sc.

6e-

55

64 .

6$

53 .

gi

se

49 6

48 .

47 .

48

45

44

figure 1

SOS 1R0 FILES FOR FIVE INSTITUTIONS

Preparatory Seal, e

Ach, A.4 As. Aft

Steubenville

Affr.

.441.4.0

Harcum I.C. cx,

Concordia cr

Delaware

Steubenville

Barry R, Morslaln

Exploratory Scales,

Mg. I.S. Inter, LA, Inv,

hrr

A MuMenberg

Delaware

Harcum IL.

Concordia

Muhlenberg

description of the ten scales. Tho five scales which dealwith the "Preparatory" orientation are:

Achievement (ach.)This scale measures the degree to which a student isoriented toward (1) the achievement of a priori goals(usually some career in particular or success ingeneral), (2) the acquisition of specific skills or cre-dentials, (3) the satisfaction of receiving external re-wards. The student who identifies with the contentsof these items has a practical, goalorientcd outlookand tends to gauge various aspects of the ;ollege ex-perience in terms of their future usefulness.Assignment Learning (A.L.)The student who agrees with a high proportion of theitems on this scale reports that ho learns best bymeeting specific, clear-cut, formal requirements. Hismode of learning Is linear, i.e., ho likes to masterspecified blocks or units of knowledge sequentially.Assessment (As.)An evaluation by those in authority seems to be quiteimportant to the student who scores high on thisscale. Grades and examinaticas are valued by thisstudent because they provide not only some measureof his abilities but some incentive for using thoseabilities.

Affiliation (Affi.)The student who prefers the manner of relating topeers expressed in items on this scale enjoys belong-ing to organized extracurricular groups. He appears tovalue the assurance of friendships such affiliationprovides. Furthermore, ho stresses the importance ofmaintaining strong institutional loyalty and support.Affirmation (Affr.)The student who agrees with items on this scale ap-pears to affirm the values of a peaceful and orderlysociety. Ho tends to support public officials in theircommitment to solve civic problems and feels "themajority can be counted on to make the right deci-sions." He would probably counsel care and cautionin the area of social change.

The five &item scales which deal with the ExploratoryOrientation are:Inquiry (inq,)

"Learning is its own reward"in essence, this is theexpressed motivation of the stude.;( who respondspositively to most of the items on this scale. Hoconcurs with statements which stress the value of in-sight, the perception of relationships, and knowing

179

EDUCATIONAL ATTITUDES AND ORIENTATIONS

how to learn, He expresses curiosity about manythings and appears to enjoy the satisfaction of inquirywhether or not It brings with it any other reward,Independent Study (La)The items on this scale help to identify the studentwho works best on his own, Ho prefers informal, un-structured coursos in which he can set his own goalsand standards and pursue his own interests, He op-pears to place a high value on freedom end inde-pendence.interaction ,infer,)An °gather' in attitude toward faculty members char-acterizes thy student with a high score on this scale.The indivic dal sea students as fully competent tosharo educational decision making with faculty. Inthis connection ho expresses the belief that studentsshould participate with faculty in planning coursesand academic programs,Informal Association (1.A.)Spontaneity marks the pattern of peer-relationshipsexpressed by the student who responds favorably tothis cluster of items, He expresses little need. foraffiliation with organized groups or for participationin formal, well planned events, His association withfellow students also tends to be unstructured,involvement (Inv.)A strong interest in social and political affairs charac-terizes the skdent who has a high score on this scaly.He sees students as having a rightful place in dealingwith the public problems of our time. Further, he ex-presses a concern for the welfare of others and stateshis readiness to take a stand on public issues.

Rest, ItriOver the course of this pilot study, undervaduates at

eight colleges and universities wore administered theStudent Orientations Survey (total N=4279)3, At this pointin the research project, the goal was to determine the de-gree to which the SOS could differentiate between studentsin various institutional settings andior curricular programs,In this regard, SOS profiles for students in five differentinstitutions are presented In standard score form in Figure1.

For these Institutions, the SOS means on the Prep-aratory scales show somewhat more variability than do themean scores on the Exploratory scares. Students at Steu-benville College (N= 149), Harcum junior College (N=92),and Concordia College (N= 718), all private institutions,tend to have higher mean scores on the Preparatory scalesof the SOS than do students at Muhlenberg College (N=425) and the University of Delaware (N=2448). With oneexception (Muhlenberg's score on Affiliation), these lattertwo institutions have scores on all Preparatory scales belowthe fiveinstitution mean,

For the Exploratory scales, there is somewhat less

180

variability in mean scores on the independent Study, Inter-action, and Involvement scales.3 Across the five insti-lutions, howover, there is moro noticeable variation inmean scores on the Inquiry and Informal Associationscales.

It was also hypothesized that the SOS profiles forstudents in different curricular arm would show as muchif not mow variability than inter-Institutional comparisons,As there were fairly largo Ns In fiom distinguishable curri-cula at the University of Delawrio, a curricular programanalysis was performed, with date from the institution. SOSprofiles for students majoring ,r1 the following areas arepresented in Figure 2: social sciences (N=433), natural sci-ences (N=317), humanities anl Brio arts (N= 317), pro -dominantly male professional curricula Engineering,Agriculture, Business (Nal 828) , and i.redominantly femaleprofessional curricula Nursing, Homo Economics. andEducation (NMI 723).

Students in the social sciences and humanities hadrelatively lower mean scores on all Preparatory scales thanstudents in the other three curricular areas. The profile forstudents in the natural sciences was slightly below thenormative mean score of 50 on these scales, and was mid-way between the profiles for students in male, professionalcurricula and those students majoring in social sciencesand humanities.

Overall, there was mote curriculum group variationin mean scores on the Exploratory scales as compared tothe interinstitutional profiles previously presented. Human-ities and social scienco majors expressed relatively moreinterest in having a participatory role with faculty ineducational decisionmaking (Interaction), had more desirein developing "learning contracts" and other independentstudy or off-campus experiences (lndep. Study), andtended to view learning as its own reward, whether or notthis learning had a practical or vocational pay-off (Inquiry).Humanities majors also had the highest mean score on theInformal Association scale (a desire for unstructured, spon-taneous peer - relationships) while social science majors, asmight be expected, had the hiihest mean score in theInvolvement scale (interest in sociopolitical issues). Onthe other hand, students in the male professional curriculahad tho lowest mean scores of all five Exploratory scales!

In a related domain, previous research on studentswho "self-select" themselves Into experimental programshad indicated that these students' general characteristicsand personality orientations are substantially differentfrom those of their peers in the regular curriculum (Heistand Bitoursky, 1971; Suzcek and Alfert, 1970). tt was hy-pothesized that differences in educational attitudes andorientations would also be evident for students in these twosettings. SOS profiles for freshmen in traditional liberalarts curricula and freshmen who voluntarily participated inexperimental programs at both the University of Hawaii

Figure 2

SOS PROFILES FOR STUDENTS IN DIFFERENT CURRICULA(UNIVERSITY OF DELAWARE)

Preparatory Scale.

Ach, AL As. Alf, Affr.

Eaptoratory Scales

Barry R, Morstain

log. I.S. Inv,

and St, Olaf's College are presented in Figure 3.At each institution, the freshmen in the experimental

program had higher mean scores on the Exploratory scalesof the SOS when compared to their peers in the regularcurriculum, Conversely, the experimental program studentshad lower mean scores than did their peers on the Prepara-tory scales, There were also substantial differences on cer-tain scales when students in the two experimental pro-grams and students in the two traditional curricula werecompared on an inter-institutional basis.

For example, freshmen in the experimental Para-College program at St. Olaf's College had higher scores onfour of five Exploratory scales of the SOS when comparedto students in the New College experimental program at theUniversity of Hawaii. In addifin, freshmen in the regularacademic curriculum at St. Ole, s College had lower meanscores on the Achievement, Affirmation, Independent

M.P. Male Professional conic.F.P. Female ProfessionalN.S. Natural SciencesS.S. Social SciencesHum.-Humanittes/Fine Arts

Study, and Interaction scales when compared to theircounterparts in the regular freshman year program atHawaii. That there would be variation of this sort is notsurprising as the two institutions are quite dissimilar withrespect to size, orientation, and admissions polities.

DiscussionBased on analyses of the data presented, the SOS op-

pears to be sensitive to the differing educational attitudesexpressed by students in various institutional settings andcurricular programs. This descriptive capability generallyrelates to the validity dimension of the SOS, as well as tothe question of how the inventory can be used by re-searchers in the field of higher education. For example, oneaspect of the impact or effectiveness of an academic pro-gram could be based in part on an analysis of the longi-tudinal changes in students' educational attitudes (Mor-

181

EDUCATIONAL ATTITUDES AND ORIENTATIONS

Bid.Sc.

117

ss

55

34

51

se

43

411

f 'Stift 3SOS PROFILES FOR FRESHMEN IN EXPERIMENTAL AND REGULAR CURRICULA

Preparatory Scales

Ad. A.L. As. At , Affr. log.

Exploratory State§

Inter. I.A. Inv.

47 Hawaii Exp, ck

40 1-

45

44

43

42

slain. 1973b).Other research studies recently completed have also

focused on the validity of the SOS. Pemberton (1973) hasshown that students with various patterns of scholastic ap-titude (as measured by SAT scores) have significantly dif-ferent SOS profiles. Morstain (1973c) administered the Om-nibus Personality Inventory and the SOS to a sample of un-dergraduates at three institutions and reported that stu-dents' educational attitudes were directly related to theirgeneral personality characteristics and orientations. Inorder to assess the relationship between individual orienta-tions and perceptions of the collegiate environment, re-search studies employing the SOS and CUES and/or theInstitutional Functioning Inventory would be of value.

182

Much discussion has also centered on the questionof the relative "fit" of student and faculty educational atti-tudes, and what bearing this may have on student develop-ment. What is the degree of congruence or incongruence ineducational altitudes and values for faculty and students,either in one class, one department, or the institution as awhole? This type of research may have implications forattempts to empirically validate hypotheses generated bythe "challenge and response" conceptualization of studentgrowth (Sanford, 1967).

From this theoretical perspective, gaining a betteroverview of student attitudes is only one side of the picture.Hence, a "Faculty Orientations Survey" has been recentlydeveloped In order to provide a means of assessing a

faculty member's educational attitudes and teachingorientations., The items in this new inventory, with appro.priate changes, correspond quite closely to items found inthe SOS. In a study which was initiated in April 1973, re-searchers at the University of Delaware are exploring thedegree of student faculty congruence in educational all!.tudes and what relationship a "disparity" factor has withrespect to how students evaluate their courses and instruc-tors. The hypothesis under consideration is that the higherthe degree of studentfaculty attitude incongruence, astudent's course/instructor ratings will be relatively towerwhen compared to the ratings of students in a "high congruence" situation.

Obviously, a boat of other research questions couldbe developed (Le. What are the educational attitudes ofstudents who change academic majors or programs? Whatare the attitudes of students who withdraw from a college.or university?). In sum, the type of research reported or

Barry R. Morstain

suggested in this paper may have implications for consider-ing an institution's academic policies and practices withreference to the question of educational outcomes. That is,many would argue that a college or university should placehigh priority on helping students assume responsibility fordeveloping the kinds of educational experiences which aremost appropriate to their particular goals, interests, andneeds. In many respects, this concern is related to White's(1952) concept of the development of a "sense of corn-petence" in this case, that a student can have some sayas to the nature of his academic experiences by taking anactive role in shaping his own education. If a more per-sonalized educational system is to occur, examining theeducation attitudes and orientations of students and facultywill of necessity assume more substantive priority in highereducation research. Hopefully, a better understanding ofleachersludent relationships and the effectiveness ofvarious educational processes will result from these efforts.

Betz, E. et. al. College student satisfaction questionnaire: monuol Ames, Iowa: Central Iowa Associates, 1971.

Cross. K.P. Beyond the open door: itew students to higher education. San Francisco: Jossey-Bass. 1971.

Feldman, K. and Newcomb. T. The impact of college on students. San Francisccr. Jossey-Bass, 1969.

Gray. R. and Morstain, B. Student orientations survey. Unpublished manuscript, 1970.

Heist, P. and Bilorusky, J. A special breed of student. In 1, Gaff (Ed.). The cluster college. San Francisco: lossey-Bass, 1970.

Heist, P. and Yonge, C. The omnibus personality inventory. New York: Psychological Corporation, 1968.

Morstain, B. Student orientations survey: manual. Newark, De.: Office of Academic Planning and Evaluation, 1973a.

Morstain, B. Changes in student's educational attitudes: a study of an experimental living learning program. Research In HigherEducation, 1973(b). 1,141-148.

Morstain, B. The relationship between students' personality characteristics and educational attitudes. (1973c). submitted toMeasurement and Evaluation in Guld.

Pace, E.R. Preliminary technico ritsonual: college and university environment scales. Princeton, N.J.: Educational Testing Service. 1963.

Pemberton. C. The relationship between student orientations to college, SAT scores, and Roteer IE scores. Research In HigherEducation (in press).

Pervin. L Satisfaction and perceived selfenvironment similarity: a semantic differential study of studentcollege interaction. Journal ofPersonality. 1967, 35. 625-634.

Peterson. R. Technical monuol: college sludenrquestionnoires. Princeton,N.J.: institutional Research Program in Higher Education,Educational Testing Service. 1965.

Sanford. N. Where colleges fall. San Francisco: Jossey-Bass. 1967.

Stern, G. Scoring instructions and college norms: activities index, college characteristics index. Syracuse. N.Y.: Psychological ResearchCenter, Syracuse University. 1963.

Suzcek. It and Alfert. E. Personality development in two different educationol atmospheres. Washington, D.C.: Department of Health.Education. and Welfare. 1970.

White, R. Lives in progress. New York: Dryden Press. 1952.

For a full discussion of the development. validity, and reliability of the Student Orientations Survey, see Morstain (1973a). Availablefrom the author.

College of Steubenville. Concordia College (Minn.), Harcum Junior College, Muhlenberg College. S. Olaf's College, University ofCalifornia at Davis, University of Delaware. University of Hawaii.

183

EDUCATIONAL ATTITUDES AND ORIENTATIONS

F values from the analysis of variance on each scale were significant at the p <01 level, with the exception of ihe Independent Studyscale ip <05).

F values from the analysis of variance on each scale were significant at the p <.01 level.

i Copies of the SOS and FOS Inventories are available frum the author,

184

COLLEGE DROPOUTS: THEORIES AND RESEARCH FINDINGS

William P. Penstemacher, Minnesota State College System

At a time when most colleges arc experiencing en-rollment decreases and concomitant financial pressures, in-creased concern is placed upon student attrition. A recentnational study by the American Council on Education indi-cated that approximately fifty-three percent (53% ) of theentering freshmen do not achieve a baccalaureate degree infour years from the institution they entered as a freshman.'Approximately forty percent of the entering freshmen infour-year colleges do not complete a baccalaureate degreewithin five years of enrolling in college.4 Colleges have be-come interested in the dropout. or the more recently de-fined stopout student, to determine ways in which the rateof attrition might be decreased.

The college -dropout has been the subject of con-siderable research during recent years. Since most empiri-cal studies of dropouts have been carried out either atsingle institutions or at the national level. with differing re-search methodologies, the findings are rarely comparableand cannot be applied to institutions in other settings. In-consistencies in the findings of attrition studies may resultfrom different populations, different institutional settings.varying definitions of dropouts, different research designs,insufficient data on the student and the college environ-ment, and a relatively high nonresponse rate. The high nor. -response rate is a common research problem in studies ofstudents who have withdrawn from college, since suchstudies are closely linked to college experiences which maybe viewed negatively. The inability to distinguish betweenstudents who withdrew with unsatisfactory academic aver-ages and those who withdrew in good academic standing isa severe limitation in interpreting the results of manystudies.

Selected Studies of AttritionThree studies of dropouts provide significant contri

buttons to the research literature for their methodology,breadth of scope, and general quality: Alexander W. Astin'sCollege Dropouts- A National Profile': a study by JeraldBachman and others, Youth in Transition DroppingOut4; and a study by Robert Cope and others entitled An4tvestisation of Entrance Characteristics Related to Typesof College Dropouts.$

College Dropouts: A National ProfileThe purpose of Astin's research was to determine

national dropout rates among full -time students attendingtwo-year colleges, and four-year colleges and universities.For students attending four-year institutions, Astin usedfour measures of persistcnce:h

Enrollment for at least a second undergraduate year;2. Receipt of a baccalaureate degree or its equivalent

within four years of entry; (47% of the studentsreceived the baccalaureate degree within four years.)

3. Continued enrollment for work toward a degree in thesame institution; (12% of the students were includedin this category.)

4. Withdrawal from the inital institution with a sub-sequent request for a transcript sent to anotherinstitution. Appri. mately twenty percent of thestudents in Astin's study were included in thiscategory.

These measures classify as dropouts only thosestudents who left their first institution-without completinga degree and who never requested transcript he sent toanother institution. Using this definition, approximatelytwenty percent (20% ) of the students in Astin's study wereclassified dropouts. When the forty-seven percent of theentering freshmen who completed a degree within fouryears is combined with the twelve percent (12% who arestill enrolled, a persistence rate of nearly sixty percent isachieved for students at four-year colleges and universities.'.While Astin found the national dropout rate to besomewhat lower than had been suggested in other studies',(McNelly": iffertq) approximately forty percent (40% ) of theentering freshmen at four-year colleges do not complete abaccalaureate degree within five years of entrance.

Youth in Trasition Dropping OutThe Youth in Transition research, a longitudinal

study of high school youth, identifies factors which predictdropping out and determines the relationship between attri-tion and changes in self-esteem, self-concepts of ability,occupational aspirations, and delinquency. Bachman constdered a commitment to education as _a broader dimension_than one which distinguished dropouts from persisters. 'rho

185

COLLEGE DROPOUTS

students who felt education was very important were notonly less likely to dropout, they were also more likely toextend their education beyond high sehoo/.10Consequenuy,Bachman conceived educational attainment as a continuumbetween high school entrance and post-high school educa-tion. The treatment of educational attainment as a con-tinuum assumes that most factors which relate to droppingout of high school also relate, in an opposite direction, toextending one's education beyond high school. In hisanalysis, three groisps of students were located on this scaleof educational attainment: at one end of the continuumwere the dropouts, at the other end were those who haveentered college or other post-high school education, and be-tween them were the non-college-bound high schoolgraduates,"

. Identifying educational attainment as a continuum isan important tool because it differentiates among highschool students who continue their education beyond highschool and those who do not. If no distinction were madebetween these two groups of persisters at the high schoollevel, it would be possible to make an interpretive error be-tween dropouts and students who stay in high school,when it is more meaningful to differentiate within the twogroups of persisting high school students.

This methodology may be applied to any institutionwishing to make distinctions within its student body. Farexample, two-year institutions could apply the method-ology to differentiate among students who do and do notenter four-year institutions. Similarly, colleges could usethis technique to distinguish between students who do anddo not continue to graduate school or other important post-graduation activity.

Before reviewing Cope's studies, it seems pertinentto discuss the use of psychological and institutional vari-ables in research on attrition. Recent studies utilizing thestandardized psychological inventories such as the Minne-sota Multiphasic Personality Inventory, the California Psy-chological Inventory, and the Omnibus Personality Inven-tory have shown that some psychological variables are re-lated to college attrition, The research has shown sub-stantial differences between dropouts and persisters andthe findings illustrate; the potential merit of consideringpsychological differenCes. A study by Rossman and Kirk atthe University of California at Berkeley differentiated be-tween "voluntary withdrawals" students with gradepoint averages of 2.0 and above and "academic failures"students with grade point averages below 2,0.12 By distinguishing between these two types of dropouts. Rossmanand Kirk found that "voluntary withdrawals" scored sig-nificantly higher than the Berkeley sophomores on two in-tellectual orientation scales, thinking Introversion and esthificism, and lower On the practital outlook scale, The

-seems of the "academic fallurei" were slightly; but not sig-nificantly lower thin the persisting students on the two in.

-186

tellectual orientation scales and slightly higher on the prac-tical outlook scale, The use of psychological tests reveal dif-ferences between types of dropouts and types of persisters,re-emphasizing that dropouts should not be considered as asingle group of students.

A growing body or research focuses on the relation-ship of the college environment, the individual's adjust-ment to college, and the process of dropping out. Severalresearchers have used the concepts of personal need andenvironmental press as a technique to study personal ad-justment in college.0 Using this approach, students areviewed as having certain needs with the strength of theseneeds characterized by one's personality. The college environment has the potential for satisfying or frustrating theneeds of the student. The term "environmental press" isapplied to the colleges potential to satisfy or thwart indi-vidual needs.

Environmental presses have been defined in termsof psychological characteristics the need for order, atolerance of ambiguity, practical orientation, intellectualorientation, and several others. When there is a discrepancybetween personal needs and the ability of the environmentto satisfy those needs, a student is more likely to dropoutthan a student who experiences a congruent relationshipbetween his needs and environmental presses. Examples ofstudies using an institutional environmental approach arethe previously cited research by Rossman and Kirk andRobert Cope.

An Investigation of Entrance Characteristics Related toTypes of College Dropouts

The work of Robert Cope and others inz...odes twostudies, one at the University of Michigan during the mid1960's and the other at the University of Washington in theearly 1970's, institutions which were found to have similarstudent characteristics at the time of the research. In thesestudies, characteristics of the college environment wore re-lated to typologies of college dropouts.

Two major presses within the college environmentwere the social and academic. A social press was definedon the basis of demographic and attitudinal informationsuch as the students' socioecoromic status, interest insocial life, politics, esthetics and other items. An academicpress was defined in terms of scores on measures of academic ability and personality orientations such as the Om-nibus Personality Inventory, The two dominant dropouttypologies included those students who dropped out forsocial reasons, and those who left for academic reasons.Examples of academically-related reasons for leaving in-chided a lack of intellectual challenge, fear of failure, oractual failure.

The research suggested that students may be incOngrunt With either or both of these major environmentalPresses in such cases, students may choose to leave 410

institution, voluntarily or involuntarily, or may remain atthe institutions, depending upon the press and the per-sonality trait being considered. Cope's data reveals that thesocialpsychological attributes of students differentiate be-tween persisters and dropouts in institutions which have Age:relatively unique social and/or academic presses.

The typologies were effective methodologically inevaluating dropouts at both the Univeristy of Michigan andthe University of Washington and may be applicable atother institutions. Cope emphasized the necessity of dif-ferentiating within the total dropout population in order torecognize the complexity of the dropout phenomenon. Forexample. the relationship between the reasons for droppingout and the characteristics of the students who dropped outemerged only after the entire group of dropouts was analyzed for within group differences and such a typology andpersonality factors.

The institutional environmental approach appears to Sex:be a promising development in the research on studentattrition. At the time the study by the Minnesota State Col-lege System was performed, psychological and institutionalenvironmental data was not available. While there has beenextensive research on attrition performed at major insiituLions such as the University of Michigan, Washington, andthe University of California at Berkeley, similar research ondropouts within state coileges has received limited Mien.lion. The research within the Minnesota State CollegeSystem may provide a comparative base for other studieswithin state colleges.

William P. Penstemacher

acteristics of dropouts and the general student body areavailable from the author.

Results of the Minnesota State College ResearchThe Minnesota State College System undertook a

study of students who had withdrawn from the six statecolleges during the fan 1971 quarter. The purpose of thestudy was to provide demographic and attitudinal informalion on the dropouts and to identify factors contributing tostudent withdrawal. Examples of the desired informationwere the student's academic background, attitudes towardtheir educational experience, reasons for withdrawing,post college activities and futLre educational plans.

The sample Included Minnesota residents enrolledas full-time students during fail 1971 who did not re enroll.for the winter 1972 quarter's Of the 2,493 Minnesota resi-dents who withdrew during or immediately after the fall1971 quarter, 2,114 or 85% were included in this study,Demographic and academic data on the 2,114 students wereprovided by the colleges Questionnaires were mailed to aneffective sample of 1,841 students. Forty percent of thequestionnaires (729 of 1,441) were returned

The following information describes the character-istics of the total group of dropouts compared with theCharacteristics of full-Bite undergraduates M th-e MinnesotaState College SyStem and the results of other attritionstudies were pertinent Additiortii'data regarding the Char.

Twenty-seven percent of the dropoutswere twenty-two years or older, which isidentical to the age distribsition of stu-dents enrolled in the Milnesota StateColleges. In other studies of dropouts,age differences were relevant onlyfor those institutions which have a significant proportion of older students. Theimportance of the age variable lies inthe increased experience; and diversifiedfamily and related demands and respon-sibilities that older students experience.

Fiftyfour percent of the dropouts weremen and forty-six percent were womenthe samr proportion of men and womenenrolled in the state colleges. Attritionstudies which have controlled for sex asan independent variable report conflitting findings, although the sex vari-able helps identify and explain dropoutbehavior when used in conjunctionwith other variables.

Marital Status: Seventeen percent of the dropouts weremarried, compared to only 12% of the undergraduate student body. Half of thosewhn were married supported dependents.Other studies have noted that marriedstudents tend to withdraw more fre-quently than unmarried students.

High SchoolRank:

ACT Scores:

Sixty percent of the dropouts were in thetop half of their high school graduatingclass. The dropouts had a slightly lowerhigh school rank than the students enrolled in the State College System. Con-sistent with other studies, there is generally a positive relationship betweenhigh school class rank and persistence incollege

The median ACT composite score for thedropouts was 20 which approximates themedian score of the students in the StateCollege System. Other research hasshown that dropouts usually have loweraptitude test sews than students continu-ing in college,

187

COLLEGE DROPOUTS

ParentalEducation:

ParentalIncome:

FinancialDependence:

Financial 'Independence:

AcademicPerformance

188

Sixty-two percent of the parents had notattended college. No comparative data forthe undergraduate student body are avail-able.

The median parental income at the timeof entering college was $7,500 whichapproximates the parental income ofthe general student body. Other studieshave indicated a trend which shows thatthe higher the socioeconomic status ofthe family, the higher the probabilityof graduating from college. It should beremembered, however, that there isan interdependent relation between education, occupation, and income. It is notany one of these factors atone which is re-lated to persistence in college: it is ratherthe combined result of education, occupa-tion, and values and, subsequently, onthe socialization of the child.

Although fewer than half of the studentswere financially dependent upon theirparents, slightly more than half of thesestudents (54% ) indicated parental in-comes of $3,000 or less. Comparable datafor the undergraduates enrolled in theMinnesota State Colleges are not avail-able.

Over half of the students who droppedout were financially self-supportingmuch higher incidence of self-supportthan occurs in the undergraduate studentbody. The financially independent stu-dents were generally older, married, andhad dependents. Over half of the self-supporting students indicated incomes of$9,000 or more

Sixty-nine percent of the dropouts hadgrade point averages of 2,0 and above atthe time if withdrawal, with a medianaverage of 2.6 after almost two years ofstudy, In contrast, the dropouts with un-satisfactory academic averages 31 per-cent of the total group of dropouts hada median grade point average of 1.5 afteronly one year of study, in comparison, 68percent of the students enrolled in the

Minnesota State College System hadcumulative grade point averages of ,0and above, while 42 percent averaged 1.9or tower. The data indicate that the drop-outs had higher academic averages thanthe currently enrolled students.

In general, the results support the findings of theAstin and Cope studies [hal lower scores on measures ofacademic ability, as defined by high school rank and testscores, and lower levels of parental education are related toattrition.

Tables 1 and 2 compare the dropouts in the Minne-sota State College System on several significant biograph-ical and academic items differentiated by sex and gradepoint average (Table I). and combinations of these vari-ables (Table 2). Students with grade point averages of 1.9 orless have been identified in the tables as "academic fail-ures," and those will.: academic averages of 2.0 as "volun-tary withdrawals."

Reasons For Dropping Out

Students rated len statements commonly given asreasons for leaving or dropping out of college as being"very important," "somewhat important," or an unim-portant factor" in their decision to withdraw. The fourreasons rated as either "very" or "somewhat important" inthe decision to withdraw were the following:

Insufficient financial resources: 48%Disappointed with the academic program: 48%Unhappy with the college experience: 47%Academic prograo not available: 38%

(Since students could indicate more than one reason as"very" or "somewhat important," the percentages exceed100% ).

A factor analysis of these four items indicates threedistinct reasons for withdrawal: the lack of financial re-sources: a general disappointment with the academicprogram combined with a general unhappiness with thecollege experience; and the unavailability of the academicprogram in which the student was interested.

However, three factors appear unique in the study ofdropouts of the Minnesota State College System:1. The importance of insufficient financial resources as a

reason for dropping out was related to a reluctance toapply for financial assistance.

2. 'rho relatively large group of dropouts who were self-supporting. Other stud_ ies of dropouts have not dis-tinguished lietween financial dependence andindependence,

3 The general disappointment with the academic pro-gram and unhappiness with the college experience.

William P. Penstemacher

Table iCOMPARISON OF BIOGRAPHICAL AND ACADEMIC VARIABLES

AMONG DROPOUTS BY SEX AND GRADE POINT AVERAGE

Biographical andAcademic Variables

SEX GRADE POINT AVERAGE

MenN=1151

WomenN=963

AcademicFailuresN= 507

VoluntaryWithdrawals

N=1134

Age > 20 yrs < 20 yrs 8.36" < 20 yrs > 20 yr 5.00"

Sex Failures were morelikely to be men 8.98"

High School Rank Third Second Third SecondQuartile Quartile 5.52" Quartile Quartile 5.72"

ACT EnglishM 16.2 18.9 10.25" 16.1 18.0 5.95"SD 5.2 5.4 5.1 5.5

ACT MathM 19.1 18.2 2.68 17.4 19.1 4.33"SD 7.0 6.7 6.6 7.1

ACT CompositeM 19.2 19,6 1.57 18.0 20.1 6.98"SD 5.5 5.2 5,0 5.5

Qtr, Firs CompetedM 74.2 71.7 0.97 47.9 87.3 13.86"SD 54.3 54.9 39.7 58.2

Satisfaction with Men were less Failures were lessAcademic Program satisfied 1.26 satisfied 3.41*

CPAM 2.06 2.46 12.05" 1.46 2.59 45.36"SD .64 4,2 .48

Note The means and standard deviations or the direction of the difference betweenmen and women, or ac.detnic failures and voluntary withdrawals are reported above.The direction of the difference is indicated if the mean of the variable was not self-explanatory. the number of respondents varied with each item.

p < .01< .001

189

COLLEGE DROPOUTS

Table 2

COMPARISON OF BIOGRAPHICAL AND ACADEMIC VARIABLESFOR ACADEMIC FAILURES AND VOLUNTARY WITHDRAWALS BY SEX

Biographical andAcademic Variables MEN WOMEN

AcademicFailuresN = 361

VoluntaryWithdrawals

N = 532

AcademicFailuresN = 146

VoluntaryWithdrawals

N = 601t

Age 20 yrs > 20 yrs. 5.67*" 19 yrs. 20 yrs 4.43"

High School Rank Third Second Third SecondQuartile Quartile 2.14 Quartile Quartile 4.31

ACT EnglishM 15.5 16.9 17.6 18.9 2.46SD 5.1 S.2 5.0 5.6

ACT MathM 18.0 19.9 3.51*" 15.8 18.5SD 6.6 7.4 6.3 6.7

ACT CompositeM 18.2 20.2 4.94." 17.4 20.0SD 5.1 S.7 4.6 3,4

Qtr. Hrs CompletedM 51.0 93.8 12.15 40.4 813 8.20*"SD 39.5 58.5 39.2 57.3

Satisfaction with Failures were less Failures were lessAcademic Program satisfied 1.96 satisfied 2.88

GPAM 1.47 2.46 34.92 1.45 2.70SD .42 .41 .42 .51

27.49

Note The means and standard deviations or the direction of the difference between theacademic failures and voluntary withdrawals for men and women are reported above.The direction of the difference is indicated if the mean of the variable was not self.explanatory. The number of respondents varied with each item.

$ p = < .05** p < .014, p < .001

100

Insufficient Financial Resources as a Reason for DroppingOut

The dropouts were separated into two groups on thebasis of financial dependence. Fifty-four percent (54%) ofthose who were dependent on parental support came fromfamilies with incomes of $3,000 or less. Within this group,only fifty-one percent (51%) applied for financial aid andseventy-six percent (76 %) received assistance. Seventy-fivepercent (75% ) of the financial aid recipients indicated thatthe amount of assistance was sufficient to meet their rea-sonable financial needs. Despite the low parental incomelevel, a relatively small percentage, only sixty -three percent(63% ), indicated the lack of financial resources as an im-portant reason for withdrawing from college. Why did sucha small percentage of students from low income familiesapply for financial aid? Among possible explanations arethe following:1. Students from tow income families percieve a low

potential for success in college and are reluctant toapply for financial assistance.

2. A strong protestant ethnic of self-support exists amongstudents from low income families. A large number ofstudents work part-time white in college. Rather thanapply for financial aid, these students would prefer towithdraw or stopout when their resources depleted,earn additional money, and then return to college.

3. An unawareness that financial aid is available.4. A reluctance to apply for or accept loans since low

income families have a history of indebtedness.One might anticipate that a large percentage of the

students who indicate financial independence have lowincomes. However, in the Minnesota State College study,only eleven percent (11 %) of the self-supporting studentshad incomes of $3,000 or less. Sixty-two percent (62% ) ofthe self-supporting students reported incomes exceeding$9,000. Only thirty-four percent (34%) of the self-support-ing students applied for financial assistance.

When the total group of dropouts is differentiatedinto groups by sex and grade point average below andabove 2.0, and then combinations of sex and grade pointaverage, the problem of financial resources is shown to berelated to certain groups of dropouts. Tables 3 and 4 compare the dropouts on the reasons for leaving college by sexand grade point average (Table 3), and combinations ofthese variables (Table 4).

the data in Tables 3 and 4 show that the lack offinancial resources was a more important problem for menthan for women, and specifically for men with satisfactoryacademic averages. The most important reason given fordropping out among meh with satisfactory grade pointaverages was a lack of financial resources. In contrast, menWith unsatisfactory academic averages ranked insufficientfinancial resources fourth, after IOW grade, unbilipinesswith the College experience, and a disapPointrnent With the

William P. Fenstemacher

academic program,The three most important reasons for women leaving

college were a disappointment with the academic program,a lack of financial resources, and an unhappiness withccllege as the three most Important reasons for droppingout.

Many of the attrition studies during the last twentyyears indicated that both men and women frequently citeddissatisfaction vith the college environment, lack of in-terest in studies, uncertain academic major and undefinedcareer plans as motives underlying the decision to with-draw. The major difference between two studies performedduring the 1950's (Iffer116, and Slocum11) and a study in the1960's (Astin and Panoo) is that the students in the laterstudy were withdrawing more for reasons of dissatisfactionwith the college environment and unsettled personal in-terests and goals. Over the past 15 or 20 years, there appearsto be an increasing tendency to cite personal factors relatedto dropping out a lack of fit between the student and theinstitutional environment, and undefined personal objec-tives. This suggests that students today, in comparison withthe students in the 1950's, are more concerned with therelevance of education to their personal growth and de-velopment.

Attitude Toward the Academic ProgramStudents who withdrew were generally satisfied

with their educational experience although they believedthe quality of instruction and the teaching-learning processcould be improved. In response to a question about theiracademic program, thirty-five percent (35%) were satisfiedwith their education, fifty-two percent (52% ) had mixedfeelings, and thirteen percent (13%) were dissatisfied withtheir education.

The quality of the faculty and course content werepositively mentioned by twenty-two percent (22%) of thestudents, yet a slightly higher percentage of students, ap-proximately twenty-seven percent (27% ), were dissatisfiedwith the instruction and course content, Forty-two percent(42% ) of the students felt their courses were not as intellectual ly challenging as they had anticipated.

Nineteen percent (19%) of the students were dis-satisfied with required courses; thirteen percent (13% )recommended elimination of general education courses orpining less emphasis on general education courses.

The results of the study suggest several curricularareas to be examined:

the purpose and the value of general educationcourses;the opportunity for students to begin concenirating In amajor field as a freshman;the need to maintain contact with students who havevoluntarily withdrawn and who are interested in con-tinning their education, through off-caMpui and

lot

COLLEGE DROPOUTS

Table 3

IMPORTANCE OF REASONS FOR WITHDRAWAL FOR DROPOUTSBY SEX AND GRADE POINT AVERAGE

MEANS AND STANDARD DEVIATIONS

Importance of SEX ' GRADE POINT AVERAGE

Reason forWithdrawal" Men

N =333WomenN c.- 388

tAcademic

FailuresN=167

VoluntaryWithdrawals

N=402t

Academic ProgramUnavailable

M 2,38 2.30 1.24 2.33 2.41 11.1SD .75 .83 .76 .77

Low GPA (#1)M 2.46 2.72 4.86" 1.93 2.85 16.90*SD .76 .S6 .76 .43

To Get MarriedM 2.90 2.63 5.68" 2.87 2.72 2.55'SD .38 .71 .45 .63

Disappointment withAcademic Program (#3) (#1) (#3) (#2)

M 2.22 2.16 0.84 2.05 2.27 2.72"SD .78 .81 .78 .79

Not Enoug Money (#1) (#3) (#1)M 2.04 2.27 3.48" 2.15 2.12 0.43SD .83 .81 .80 .8S

Health ReasonsM 2.81 2.70 2.36" 2.77 2.78 0.21SD .51 .67 .56 .57

EnlisteraftedM 2.68 2.92 5.32"' 2.78 2.82 0.63SD .70 .37 .60 .55

Unhappy withCollege Experience (#2) (#2) (#2) (#3)

M 2.15 2.21 0.95 1.99 2.29 3.68600

SD .83 .83 .84 .80Family Responsibilities

M 2.73 2.58 2.72" 2.68 2.65 0.47SO . .60 .72 .65 .68

Unsatisfactory HousingM 2.68 2.64 0.71 2.67 2.67 0.12SD .61 .64 .63 .63

Note - The numbers in parentheses indicate the rank among the three most importantreasons for withdrawal within each group.

The code ranged from I to 3; 1 = Reason was rated "very Important;" 2 = "somewhatInspcolariV 3 = "not at all important" in the derision to withdraw. the number ofrespondents varied with each variable.

p < .05#* p < .01

44`p < .001

192

William P. Fonstemacher

Table 4

IMPORTANCE OF REASONS FOR WITHDRAWALS FORACADEMIC FAILURES AND VOLUNTARY WITHDRAWALS BY SEX

MEANS AND STANDARD DEVIATIONS

MEN WOMENImportance ofReason forWithdrawal '

AcademicFailuresN=116

VoluntaryWithdrawals

N 157t

AcademicFailuresN= 51

VoluntaryWithdrawals

N= 244I

Academic ProgramUnavailable

M 2.41 2.46 2.15 2.38 1.70SD .71 .72

0.55.82 .80

Low GPA (#1) (#3)M 1.89 2.78 10.09" 2.00 2.89 13.12"SD .79 .55 .67 .33

To Get MarriedM 2.92 2.88 0.85 2.76 2.62 1.15SD .34 .43 .62 0 .72

Disappointment withAcademic Program (#3) (#3) (#2) (#1)

M 2.13 2.33 1.98 1,88 2.23 2.63SD .79 .76 .7S .80 -

Not Enough Money (#1) (#2)M 2,15 1.94 1.93 2,16 2.24 0.54SD .79 .85 .84 .83

Hralth ReasonsM 2.75 2.86 1,70 2,82 2.73 0.84SD .59 .42 .51 .64

Enlisted or DraftedM 2 .69 2. 64 0.54 3.00 2.94 1.18SD .69 .74 .00 .33

Unhappy withCollege Experience (#2) (#2) (#1) (#3)

M 2,07 2.25 1.58 1.82 2.33 3.86**SD .86 .80 .78 .80

Family Responsibilitiesa M 2.69 2.75 0.78 2.66 2.58 0.60

SD .67 .55 .63 .75Unsatisfactory Housing

M 2.67 2.70 0.42 2,67 2.66 0.10SO .63 .61 .62 .64

Note - the numbers In parentheses indicate the rank among the three most Importantreasons for withdrawal within each group.

The code ranged from 1 to 3; 1= Reason was rated "very Important ;" 2= "somewhatIMportatit;"--3="not at all important" in the decision to withdraw. the number ofrespondents varied with each variable.

* p < .010 tt. < 401

103

COLLEGE DROPOUTS

external studies programs.The purpose and value of general education and the

liberal arts need to be examined in the context of the ques-tions which students are facing today, questions involvingthe role of the liberal arts among a pragmatically orientedstudent body. Rather than require courses of all studentswithin the first two years, an alternative might be an ar-rangement which would allow students to demonstrate acompetency in the liberal arts prior to graduation. Thiswould enable students who enter college Intent on com-pleting a major tho opportunity to take the courses in whichthey are interested rather than postpone the courses for twoyears. If the student were allowed, with the guidance offaculty, to design a series of experiences in the liberal arts.he should be able to develop a program of study compli-mentary to a major area of study, an intended vocation, andlarger societal concerns.

By eliminating the general education or liberal artsrequirements, faculty might have the freedom to designcourses which may have greater appeal to students, Anappreciation of the liberal arts cannot be attained throughregulations. The college environment the student,faculty, in and outof class activities should complimentand encourage an interest in and appreciation of the arts.

The vast majority of students who dropped out of theMinnesota State Colleges plan to continue their post-secondary education. Approximately forty percent (40% )were enrolled in a college or vocational school during theimmediate next quarter, and eighty percent (80%) intend toenroll in a school or college during the next year. With sucha large number of students planning to continue theireducation, we are dealing with a stopout phenomenon, onein which students temporarily withdraw from the formalpost-secondary program to participate in more meaningfulactivities. For students who cannot continue as on-campusstudents, external studies arrangements provide an oppor-tunity for these individuals to continue to study and main-tain contact with the faculty and the college.

SummaryThis paper has briefly reviewed selected research on

college dropouts and discussed the findings of several attri-tion studies Dropping out of a college is a very complexprocess and comprehension of the problem requires asubstantial amount of information about the student andthe institution, The diversity within the dropout groupindicates that dropouts should not be grouped into onecategory for analysis since the significant factors related tothe attrition of one group of dropouts will very likely bemasked by the total group. At a minimum, distinctionsshould be made on glade point average, sex, and

194

combinations of these variables.In addition to the importance of biographical, aca-

demic, and psychological information on students, the re-search involving the relationship between the student andthe college environment appears likely to have the mostrelevance in increasing our understanding of the dropoutphenomenon. The student-environmental fit approachsimultaneously studies the effects of several biographical,educational, psychological, and environmental variables.To implement this approach, colleges will need to obtainpsychological data on the student body, from whichinformation about environmental press may be derived.

Several findings from the Minnesota State Collegeresearch have hopefully, contributed to the information ondropouts.

Information regarding the dependence-independenceof parental financial support appears to be important.The extent of self-support among the stopouls in theMinnesota State Colleges was higher than anticipated.Financial difficulty, or the lack of financial resources,was among the three or four most frequently citedreasons for withdrawal. However, when the total groupof dropouts was differentialid by sex and academiCaverage, the lack of financial resources was thedominant reason for withdrawal of only one groupmen with satIsfactory academic averages.Students with unsatisfactory academic averages tend toview their college experience in negative ways. Whilethey perhaps correctly attribute their primary reasonfor leaving college as an unsatisfactory academicaverage, the worry and fear of failure apparently spillsover into an unhappiness with the college experienceand a disappointment with the academic program. Alack of financial resources was ranked fourth bystudents with unsatisfactory averages, so we canassume that they were not primarily worried abouttheir financial ability to remain in college.Information regarding student altitudes toward theacademic program provides a means of assessingconsumer response to the academic fare, The educa-tional experiences identified as significantly positiveor negative, and the reasons for these comments,should be explored to identify possible areas of curri-cular or instructional change, For example, there maybe areas of the academic program which students withcertain important characteristics, such as the voluntarywithdrawals. believe the college is doing well, andareas which might be improved. The findings may thenbe used to support alternative cuiricular programsdesigned to more adequately satisfy student needs

William P. Fenstemacher

Alexander W. Astir), College Dropouts: A National Profile, Research Report, Vol. 7, No. 1. Washington: American Council onEducation, 1972, p. 10.

+Ibid., p. 49.

Ilbid.

4ferald G. Bachman, Swayzer Green, and Ilona D. Wirlantm, Youth in Transition - Dropping Oul. Vol. III. Ann Arbor, Michigan:Institute for Social Research, 1971,

sRobeil G. Cope. Raymond G. Hewitt, Keith G. Pailthorp. Michael Skating, and David C. Trapp. An Investigation of EntranceCharacteristics Related to Types of College Dropouts. Washington: U.S. Office of Education, 1971.

" Astin, p. 4.

p. 10.

9.11. McNelly, College Student Mortality. Washington : U.S. Office of Education, Bulletin 1937, No. 11. 1938.

vRobert E. Wert. Retention and Withdrawal of College Students. Washington: U.S. Office of Education, 1957, Series 1958, No. 1.

1013achman and others, p. 7.

p. 8.

niack E. Rossmann and Barbara A. Kirk, "Factors Related to Persistence and Withdrawal among University Students." Journal ofCounseling Psychology, XVII, No. 1, pp. 56-62.

$iSeveral original sources describing the concepts of personal need and environmental press include:H.A. Murray, Explorations in Personality. New York: Oxford University Press, 19:38.C.R. Pace, "Implications of Differences in Campus Atmosphere for Evaluation and Planning of College Programs." R.L. Sutherland el al.

(Eds.),Personality Factors on The College Campus: Review of a Symposium. Austin. Texas: The Hogg Foundation for Mental Health,University of Texas. 1962, pp. 43-61.

C.R. Pace and C.C. Stern, ''An Approach to the Measurement of Psychological Characteristics of College Environments," Journal ofEducational Psychology,IL, pp. 269-277.

G.G. Stern, "Environments for Learning," N. Sanford (Ed,). The American College: A Psychological and Social Interprelotion of theHigher Learning. New York: Wiley, 1962, pp. 690-730.

ilCope and others. pp. 84-94.

'SIn the Minnesota State College study. a dropout was defined as a stale resident enrolled full -time during fall 1971 who did not re-enrollthe following terms as a full- or part-time student. This definition includes as dropouts the students with both satisfactory and unsatisfactoryacademic averages, although a distinction was made between these two groups:those with satisfactory academic averages were classified as"voluntary dropouts:" those with unsatisfactory averages as "involuntary dropouts." Some researchers have noted the imprecision of the term"dropout," Students who voluntarily withdraw are generally considered dropouts. although a number of these students may transfer to anotherinstitution and/or subsequently return to the initial institution. The term "stapout" has recently been used to refer to students who temporarilyleave post-secondary education and return at a later time.

solfferl.

Slocum. Academic Mortality at the State College of Washington, Pullman: State College of Washington, 1956.

18Alexander W. Astin and Robert j. Panos. The Educational and Vocational Development of College Students. Washington: AmericanCouncil on Education, 1969.

195