unraveling the mystery of institutionalization

7
Unraveling the Mystery of Institutionalization ;MIA'-I'HEW B. MILES "The remedial labs? 'Ihes woiuld keep them going unless something realls wrong happened. T'hat's the was it is around here. Things keep going as long as there aren't waves"--I ab teacher, Banestow n site.' "' eachers get no rewards for using neu programs. XiXe do neu things all the time, so the new legal program, IPL:. isn't special And besides, unless sie implement IPLE as as a hole program. not just pieces of it, teachers will resist it because of all the handout copying we have to do now. We don't have a clerical staff to do that. "--''cacher, Burton site. 'All site names are pseudonsms Whether or not a program becomes a durable part of the curriculum depends on teacher mastery and commitment and administrative action as well as other factors. hese two schools "-crc part of ao inteisise study of 12 cclmnicntarx and secondary sclhools included in the Study of Disscminatioin IEfforts Supporting School Improscnicilt iCranl- dall and associates, 1982: Hubcrnlan and \liles, 19821 In both schools. the iiiosatiou ~sas being carried out effcctisclx. \Yet a ycar later, oe schiool's program, like itllan) "ness things." had disappeared. TIhe other had conitinrued, by soiieho;N get- ting "built ili" to the life of the school Why does this institutiolnalization oc- cur? Illntil rcccittl, there has been ecry little rcsearc h done on this qucstion. We are faeed vsith a mystery, %sith fcss reliable guidelines to help us unrascl it Most work on how. schools change has stressed the "front end" of the proccss- hou inniovations arc adlopted in the first place (Rogers, 1962; Zaltman, Duncan, and Holbek. 1973). There has beeii a grow-ing body of swork over the past fes years oil issues of implerentation f'lul- lan, 1982). but it has not been centrally focused on institutionalization. At best, researchers examiined "continuatior"; for instance, a news reading program might well be continued if, say, a build- ing principal or a key teacher liked it or found it nmore convenient than somie other practice. But what happens if the key advocate leaves? Without sonic sense of "built-in-ncss," the fate of in- novations is in doubt. ''he work of Berman and McLaughlin ( 1977) clearl suggests that federally funded programs tend not to be continued after the funds terminate. In short, as Huberman and Crandall (1983) remark: In the chronicle of research on dissemina- tion and use of educational practices, we first put our chips on adoption, then on imple- mentation It turns out that these invest- ments are lost without deliberate attention to the institutional steps that lock an innovation into tlhe local settinlg \cs pI)racticc, thlalt get built iln to thc training, rcgtditorr stffiatng anid blidgc'tar% cl c Ir' niIc: othiIcr' (1 In't linnloatins Irc highlll picrs lihlc giod'. I akig stitlitittiiiahiatioll for grailteI -II s- slIlrnIig socll i hait l.ingit all tlhat it N,1 ilp- pcln h itsielf.t or sill neccssaril rcsilt froii i tcclniicalls imaistcred dciionistrail)l ctfcc tc pronlcct-is ilisc a.ld iisuall sclf-idcfc.lting 'I'lhere lias bee] sollic scattered(l ssirk oni instituitiou alizatiion in the past fcsi vcicrs 'Y i and others (198), fior i- stanice, studiced institutioializatioll is- sues in public orgalizatioiis. I lc (Ic- -clopcd a thoughtful foilmulatio in ordler to get blilt ill or routilized, iil(no- vationls nmust Ihac slillillillu su)port hb 14 LncicATioNAL Li,AnFRsHip 14 F.DUCATIONAI. LEADFRSHIP

Upload: dangdan

Post on 01-Jan-2017

234 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: Unraveling the Mystery of Institutionalization

Unraveling the Mystery ofInstitutionalization

;MIA'-I'HEW B. MILES

"The remedial labs? 'Ihes woiuld keepthem going unless something reallswrong happened. T'hat's the was it isaround here. Things keep going as longas there aren't waves"--I ab teacher,Banestow n site.'

"' eachers get no rewards for using neuprograms. XiXe do neu things all thetime, so the new legal program, IPL:.isn't special And besides, unless sieimplement IPLE as as a hole program.not just pieces of it, teachers will resist itbecause of all the handout copying wehave to do now. We don't have a clericalstaff to do that. "--''cacher, Burton site.'All site names are pseudonsms

Whether or not aprogram becomesa durable part ofthe curriculumdepends onteacher masteryand commitmentand administrativeaction as well asother factors.

hese two schools "-crc part of aointeisise study of 12 cclmnicntarxand secondary sclhools included

in the Study of Disscminatioin IEfforts

Supporting School Improscnicilt iCranl-dall and associates, 1982: Hubcrnlanand \liles, 19821

In both schools. the iiiosatiou ~sasbeing carried out effcctisclx. \Yet a ycarlater, oe schiool's program, like itllan)

"ness things." had disappeared. TIheother had conitinrued, by soiieho;N get-ting "built ili" to the life of the schoolWhy does this institutiolnalization oc-cur?

Illntil rcccittl, there has been ecrylittle rcsearc h done on this qucstion.We are faeed vsith a mystery, %sith fcssreliable guidelines to help us unrascl itMost work on how. schools change hasstressed the "front end" of the proccss-hou inniovations arc adlopted in the firstplace (Rogers, 1962; Zaltman, Duncan,and Holbek. 1973). There has beeii agrow-ing body of swork over the past fesyears oil issues of implerentation f'lul-lan, 1982). but it has not been centrallyfocused on institutionalization. At best,researchers examiined "continuatior";for instance, a news reading programmight well be continued if, say, a build-ing principal or a key teacher liked it orfound it nmore convenient than somieother practice. But what happens if thekey advocate leaves? Without sonicsense of "built-in-ncss," the fate of in-novations is in doubt. ''he work ofBerman and McLaughlin ( 1977) clearlsuggests that federally funded programstend not to be continued after the fundsterminate. In short, as Huberman andCrandall (1983) remark:

In the chronicle of research on dissemina-tion and use of educational practices, we firstput our chips on adoption, then on imple-mentation It turns out that these invest-ments are lost without deliberate attention tothe institutional steps that lock an innovation

into tlhe local settinlg \cs pI)racticc, thlalt getbuilt iln to thc training, rcgtditorr stffiatnganid blidgc'tar% cl c Ir' niIc: othiIcr' (1 In'tlinnloatins Irc highlll picrs lihlc giod'.I akig stitlitittiiiahiatioll for grailteI -II s-

slIlrnIig socll i hait l.ingit all tlhat it N,1 ilp-pcln h itsielf.t or sill neccssaril rcsilt froii itcclniicalls imaistcred dciionistrail)l ctfcc tcpronlcct-is ilisc a.ld iisuall sclf-idcfc.lting

'I'lhere lias bee] sollic scattered(l ssirkoni instituitiou alizatiion in the past fcsivcicrs 'Y i and others (198), fior i-stanice, studiced institutioializatioll is-sues in public orgalizatioiis. I lc (Ic--clopcd a thoughtful foilmulatio inordler to get blilt ill or routilized, iil(no-vationls nmust Ihac slillillillu su)port hb

14 LncicATioNAL Li�,AnFRsHip

14 F.DUCATIONAI. LEADFRSHIP

Page 2: Unraveling the Mystery of Institutionalization

risers alld Ilialagcrs. Ilnust COniplete cer-

tain "passages" (for example. going fromsoft to hard nlioncs), and nrist sunviveccrtail "csclcs" (Incs budget rounds orpersonnel turnosmcr) But their studieswcerc limited to "tcchnological" inlloa-tions, such as closed-circuit tclexision,and incldel d onl- a fec school districtcases along s-ith tlheir studies of fire andpolice departlmentls Wc don't knowwhether their findings sould apply toinnov.atiolis across a %vidc range ofschools and school districts.

Leakc and othcrs (1978) dc clopedsome I ers interesting training materialson institutionalizationll for the ITeacher

Corps. treating it as a strategic processcontaining mans specific milcstoncsand critical events, wvith associated strat-cgics, such as collaboratievc involhc-mcnt. adoministrative fiat. and "owncr-ship enlargement and transfer." lhework also usefullh stressed the impor-tancc of political aspects of institutional-iZation and attention to thce actions ofkc gatekeepers. While the Leakc mate-rials. s\hich w ere designed as part of ageneral training package for peoplefaced wsith institutionalization issues,arc plausible and probably helpfiul topeople in schools. thes\ lack a researchbase.

Corbett. l)a-son. and Firestonc(1982) studied 14 schools working with aregional lab on problems of basic skillsand career and citizen education, anddeveloped a general model of w-hat ledto continuation of the innovations beingattempted. Thcs usefull' focused onpost-implementation cecnts such as pro-vision of incentives and assessment ofeffectiveness. but had little data onbroader structural changes that might benceded in schools to ensure continueduse. Thc% noted that even the officialincorporation of an innovation into theschool's cumculum did not guaranteedurable continuation. Similarld. Gla-scr's ( 1981) review' of durability of inno-vations in human snrvice organizationstended to stress staff opportunities todiscuss implemented changes and re-ceive feedback and reinforcement. Thepnrimary focus was on the user's "person-al involvemnlent' organizational levelchanges were treated as a vague back-drop.

The involvement theme also ap-peared in Louis and others' i 1981) studhof the R&D Utilization Program (90schools), though effects of pre-cxistingsite characteristics and of the innoatiabonitself had strong influence as well. TheLouis stud! also found that "incorpo-ration" of the problem-solving processused in the program waas less frequent(and less predictable) than incorporationof the specific innovahive products in-volved.

Howes (1977) proposed a general"contingency" model for predicting in-stitutionalization. and tested it on asample of data from eight schools andthree correctional facilities. Her find-ings stressed the importance of support-

Matthew B. Miles is Senior ResearchAssociate, Center for Polres Research.New York. New, York

NOVEMBER 1983 I1

)pportunities to discuss changes and receive fe

Page 3: Unraveling the Mystery of Institutionalization

"For each of the 12sites, we drew a 'causalnetwork' that describedthe course of eventsduring implementation."

ive informal networks, full commurtica-tion, coordination, and technicalsupport. But her results were based onlyon a survey "snapshot" and did not lookat actual institutionalization decisionsand dynamics over time. Thus she con-cluded that a "contingency" approachwas required; the demands of each inno-vation and school settings v'erc likely tobe different. Berman (19811 echoes thiscontingency approach.

In short, past research and conxen-tional wisdom tend to suggest that a"good," well-mastered innovation thatits users endorse or support will somre-how just stay around. There has beenoveremphasis on user ownership, in-volvement, and technical skill; the orga-nization-level structural and proceduralchanges required for institutionalizationhave staved vague and mysterious. 'i-nallv, the empirical data base on hbuvinstitutionalization actually works overtime is quite thin.

The rest of this article describes re-cent research that begins to rernedythese uncertainties and takes steps to-ward unraveling the mystery.

Studying InstitutionalizationIn the field study component of theStudy of Dissemination Efforts Support-ing School Improvement, wes exam-ined improvement processes in 12schools in rural, suburban, and urbansettings in ten states. Adapting classicethnographic methods, we collecteddata through nonparticipant observa-tion, semi-structured and informal in-terviews, and documents during three tofour intensive site visits over the 1979-80 school year. Our research questionsdealt with many aspects of the schoolimprovement process: the innovationsthemselves; the local context; the assist-ance provided; "transformations" in theinnovation, the user, and the school;and the intermediate and final out-

Figure 1. Institutionalization of a Remedial Lab, Banestown Site

School Level District Level

Supporting Conditions:Is a core (vg. peripheral) application

Operating on regular, daily basis

Provides benefits, payoffs to users

Competing practices eliminatedReceives support from:

Administrators

Users/staffClients

Passage Completion:Goes from soft to hard moneylob description becomes standardSkills required are included in formal training

programOrganizational status is established/part of

regulationsRoutines established for supply and

maintenanceCycle Survival:Survives annual budget cyclesSurvives departure or introduction of new

personnelSkills are taught in successive cycles

Achieves widespread use throughoutorganization

Survives equipment turnover or loss (includesmaterials)

comes of the effort, inclndinrg institu-tionalization.

The 12 sites included sesen that screimplementing National Diffusion Net-work (NDN) innovations ranging frominreading and math programs to careereducation, environmental studies. earlschildhood education, and govcrnmen-tal/legal issues. There were also fiveTitle IV-C sites that had developed pro-grams in social studies, individualizcdeducation. reading, vocational cdua-tion, and a complete alternative school

We analyzed our 2,700 pages of fieldnotes using a common format of tables,charts, and narrative text, which result-ed in 12 case reports. We then wrote alengthy cross-site analysis (f lubermliaand Miles, 1982) using multi-site rtatri-ces and causal networks to ldevelop gen-eralizations and explanations that madesense across the 12 sites while respectingthe unique aspects of each, 2

At field sites, we looked for organiza-tional conditions supporting institution-alization, asking people not only wheth-er the innovation would be around inthe following and subsequent years, butwhy they thought so. In addition, eachresearcher filled out a standard checklist

present,temporarypresent,

temporarypresent, looks

dubiousdubious, weak

present

presentpresent

absentdubious

absent

absentpresent,

temporary

absentpromised

present,temporary

present,temporary

n/a

promised

present,temporary

present, lookspermanent

dubious, weak

present, partiallysoft

presentpresent

absentdubious

absent

absentpresent,

temporary

absentpromised

present.temporary

present,temporary

n/a

chart. Fignrc I is a sample chart shions-ing the moderate degree of irstitultional-ization at the Banestosin site. The con-ceptualizationi unlclerlsing the chart isdrawn from Yili and others (1978)

'Ihe analyst also comnentted on arldarmplified the chart with accompatxsitigtext For example, at Bancstosin, theanalyst noted that though the remedliallab had proved to be a better solution toneeds than the existing practices ('itle Iaides or in-class s ork). and thus hadgood support from teachers and thecentral office'. other signs were not goodl.For instance, the district budget line forthe lab was still dravn from soft uritncc;supervisory responsibilith for the pro-gram had shifted and might shift againiand the lab did not have a "firm institiu-tional status." Still, the analyst wasmore optimistic than the chart suggests.emphasizing central office administra-tors' and users' strong belief in the needfor the project. He noted the existenceof similar labs at four other schools inthe county ("cutting off the lab at theelementary school would jeopardize theinstitutional rationale at the middleschool"), and cited the lab teacher'scomment that "things keep going as

16 I�.oueATioNAc Li'.An�.RsHip

l

16 EDUCATIONAL LEAI)DERSHIP

Page 4: Unraveling the Mystery of Institutionalization

long as there arei't big uwacs."Finally. the analyst pointed out that

recent budget cuts had ot elinminatedthe lab at another school, and that thetwo lab teachers alrcads on the counhrpayroll would almrost surcly not Ix fired.1The text concludes:IThle most likcls prcdiction is thai lhe labs arc

high oil tile list of proiects to be supportedonce core c lassro0ri1 practices ha-ce beenassurcd T1hc labs would lc sacrificed bcforcfunds for tcac hers or core equlipllenlt orinstructillnal materials Bec ond that. the labsarc high oil thce list of priorities for "speci~alprograms

We begin to see here the comnplcxitof tceasing out the factors that "ill let usmake a confident pred(icti(io n

Natrrally. other sites differed in houswell institutionalized their imtloatiollsseemed to be. In 'Tindalc. a high-insti-tutionalization site, a localls dcclopedreading program got researcher ratingsof "present. looks pernlmancnt" on almolstevyer aspect of our chart Thic ncu-programr "replaced the old lEnglishl cur-riculuni and funds likc an!s other pro-gram in .e.c school" All users sawrs it as"built inm," an(l "sonic had ncever knosniiit not to Ix'."

T'Ihe ultimatc measuire of institution-alization at T'lindalc. ironicall enough,appears in this sentlncc: "Ill severalyears, it will he recised. just as all other

curricula arc rcvised on a regular basis."F'or another example, institutional-

ization was essentialls absent at theBurton site. Thec central office socialstudies coordinator defiued the IPLE.legal education unit as "experimental"and gave teachers a license to "pick andchoose" at will from the IPILE: materials.In practicalls all cells of the chart. theanalyst gave a rating of "absent." On)Ione uscr of four was using the materialsfairly rcgularly. and said thc impirumcdteaching. A fcus students said thcs likedthe cxercises, and thci social studiescoordinator w-as supportiec.

But thce chances for institlitionaliza-tion looked slim. The cominillrir -basedand other experiential actisitics of theprogram had a crn poor fit with users'thpical practices. \W hetlhcr the IPL.I'materials would ever Ix institution-alized del'nded on curricuilurlm rc-icsscommittees O()ne user noted. "If therevision work is done bs ain IPLE sup-porter or two. therl IPLE would get builtil. If it's not done bs a sulpportcr. IP.I'.would be used less. much Icess. T''cachcrsarc free to use as mulch of a rcxiscdcurriculum as thc\ \ant or lecd." Andanothcr riser, as sue sass at the start ofthis article. noted lack of rc\ards andteacher resistance.

All in all. there uas picnth of saria-

tion across our 12 sites. Three had highinstitutionalizatiol. three had high tomoderate. one had moderate. three hadlou, and tho could onls get a rathng of"nil." 'hirs range enabled us to lookcarcfillh at predictine factors.

Explaining InstitutionalizationFor each of thle 12 sites. we drcs a"causal rtchwork" that descrinbed tKcourse of esicuts durinlg inlplcrmcutahonleading to high. molderate. or lom out-comes. includiing instihtionalizatioll.(\Ve fed thes nCretiorks back to peOpk atthe sites. swho confimied their essentialsalidith. as rell as refining them fur-ther. ) \Ve founld that wre ere able tosort the 12 sites into four differntl "fam-ilics." according to their approach toinstitutionalizathiol issucs: htso fanliliesuWere high and hto hlos

Ii1 our first scenario. we foiund olK'famrril of four sites that solhed institu-tionaliiationl issucs through mandated.stable use-that is. requiring the use tfthe ininov ation. and maintaininlg muchstabilith in personnel and their use ofthe imlovation' I'gure 2 reprcscnts 'lin-dale's ston.

Hecrc, a posscrfild ccntral (fficeadministrator I I ). tile director of cumc-rlum and special proic'ts. suorkitg froma centralized porwer base t21. put consid-

NOVFMB�.R 1983 I-

Figure 2. Mandated, Stable Use as a Route to Institutionalization, rmndle Site

+ Influence of other variables not discussed.

NOVFMBER 1983 I-

Page 5: Unraveling the Mystery of Institutionalization

erable pressure on users (31 to implc-ment the nciu. locall! de-eloped read-ing program. Initialls. this loveredusers' commitment 34j thes resentedand feared the pressure. But substantialassistance (5) sas supplied, vhich in-creased users' practice mastcr t6) agood deal and subsequenctly their conl-mitment (4) In addition, organizationalrearrangements (%7. including schIcdil-ing, pupil rotation. and teacher teamingssere made, inicrasing student impact(8i. User masters andl commitmenrt.along s ith stabilits of programl staff i9i,led to stabilized use ll)3. x hich bothincreased percentage of usc ( I 11 ) and ledto institutionalizatioll (12). Stability ofprogram leadership i 13) also aided insti-tutionalization. The general picture isone of administrati\Ce dlcisi eness ac-companied bh enough assistance to in-crease user skill, onriership. anld stableuse in the context of a stable sssterli

A second institutionlalization scenar-io. Ahich got moderate to high out-comes, was skillful, committed use. Thisfamih of three sites did not mandate theinnovatioi, but spent much energ- o3nassisting users and deseloping theircommitment.

:\ third, less effecti\e sccnario. ap-pearing in hto sites. scs called vulnera-bility; w hen fundinrg crises struck or kc\advocates left the district, the imro a-tion had no protection Ksxcn thorighlusers verc rcasonablk skilled andr corm-

nmitted. there \as little guaraltce ordurabilith.

Fi nalls. \e sas ali eCssenttiall noil nll-stitutionializing sCCelari o ifi three sites.awhich *'e labeled indifference Adiiuiis-trators, in particular. shricd bh\ theirbcha\ior that the'\ did not care. h'liedid almost nothinlg to assist users ill airl.as

A General ModelLooking at these four sccnarios. *cextracted a list of 2i3 kcs \a-riables thatseemed to be irirol\ed. examining thenmin a sites-bsh\-arialles, prcdlictor-out-corre matrix to see * bo each contribult-ed to high or los\ results ITbhi ceassembled the most crucial into a gienr-al model. hich appears in l'igure 3.()ir ermpirical studs of the 12 sitesshotlecd ts. im brief, that institnitioiial-ization mulst be approachedl b! pro\ id-ing supports anll by wsarding off threats

I[he story begins in the uipper leftcrrner xuith administrative commitment1 ). That's a necessary but, as a\c shall

see. insufficicnt cornditionr for high iisti-tutionalization Our analysis of the see-narios suggests clearls that high adnlin-istratise corinilitnient tends to lead toboth administrative pressure (2! ol usersto implemcelt tile inrlo\atioll o. g111

wsith administrative support i) . Ihicihoften sho-ss up ini the formr of assistance(6) to users. Both the pressure adl theassistance tend to Icad] to increasedt user

effort (5). e rcpcatells fiirilll that theliarder people xrorked at all ininilvation.thIe more committed (9)) thce grea; thlatcorrim tillent sas also fuieled bs increas-ing technlical master of thie illo\xatio

I1))Cominliitniicnt anid imasters both lea(l

toraard increasinrg stabhilizatin of use(8); the irlnovation has "settled dlon" inthe sstem. I hat stabilizaition is alsoaided if ladminlistrators decide to maln-date (4t the irlos\atiolr. lticl also ilat-

urall\ inicreases the percentage of use (3-to soiiiethig approachmiig Ili0) percentof eligible users; that iin itself dccisi\chlencourages institutiona';zation ( 12iBrit here is one mnore critical factorWhere adminuistratrs scrc committed(I , the\ also took direct action to bringabout organizational change (I I-chIaniges Ibeond thlose tile stabilizcld ill-nov ationi had alreads broiight. In partic-ular, tles aorkcd at the "passages" ainl"cscls" shoiun ill our charts bh altering

the structure anld approalclh of inscr vietraininlg. ariting the Imno\ation's rc-

quiremcn ts into job dcescriptions. all1k-

ing nc\s bludlget lilies. aIppoinlting per-niaucnt coordlirators for thic inloi atioll.and makinig Sure tlhat ieeded matcrialsalnd clqulipuenlit sorloul corntinille to bea\ailable in slcceedinlg sears.

All these supports for institutionalliza-tion rrladC empirical sense ill our sitesBut the lessoni of our low-institiltifiializ-ing sites is that positi\e supports are lnot

18 EDCII(ATI()ONAL. L1EADERSHIP

Figure 3. A Data-Grounded Model of Institutionalization

4

Page 6: Unraveling the Mystery of Institutionalization

enough. It's neccssar to ward off threatsto the durability of the innovation. Inour sites, these threats arose fromi t osources. First. there swas environmentalturbulence (1 ), usIlall- in the form offunding cuts or losses, but sometimes inthe form of shifting or shrinking studentpopulations. Second. we sasw career ad-vancement motivation (14), the genuinedesire of professionals to nmoxe on tonew challenges. Both served as threatsto institutionlalizationll becausc they de-stabilized both program staff (li) andleadership l(6) As one superintendentmused when a kes principal took a jobin anotler district to promote the \ecninnovatiomi he had advocated and sup-ported locall. "It's a temporarn setback.but wc'll just keep going. It's a good dealfor Bill. It helps him ii his growth to bemoing oin. A loss? ()f course it's a loss."And the innoS ation's coordinator. s hoherself wsas also mos ing to a nesc job.said of the superinted cndt. "You has-e toappreciate the energy hIe's put into it. Itmust bhe tough to have to be doing it allover again.

So job imobilit- hethcr drixen b-advanccient motivatiorl or bs fundingcuts, is a threat to institutionalization.The innovation miust be buffered, pro-tected against these threats. or it w-illbecome highl! vulnerable (1-) Onceagain, as our model sho\\s, organiza-tional change I I I) is critical. If stnlctur-al and procedural changes have oc-curred. \ ulllcrabilit is reduced. Forexample. in the site sec just dres quotesfrom, the decisioll to create a dis-trichlidc managecmcnt group for the in-novation. inrolhing both teac hers andadministrators. nlrtured stabliliht e\enthough the principal and the coordina-tor were both leaving. An external con-sultant noted. "'Ihe prograll has abroad foundation in the malnagementteamn and has good feedback from par-ents. Tlhey w on't go back to wshat theswere doing before." And a teacher u'howas an earl! program advocate said. "Ifthe program 'were being pushed by oneperson. uwhen that person left. it wouldbe dead. But if seC en or eight people aremaking decisions, the likelihood of stir-viving personnel changes is great."

The general message of our modxel isthat the enthusiasml, skill. and effective-ness of the inlovation are insufficientconditions for institutionalization.Rather, what seems required is strongattention of admillistrators to stabilizingand supportilig the inmloation. exterld-ing its use to a large group. and makingprovisions to protect the innovation

against the threats of pcrsolinel tirnmoicrthat are enidemi in schools. Nlakingclearcut changes in organizationalstructlre. rules. and procedures seceCmsessential both to stabilize the innlovationand to buffer against turnover.

EpilogueWhen we completed this nmodel. cueapplied it to each of our 12 sites. basedon all the data available. and made aspecific prediction about the degree ofinstitutionalizatioi that swould be prcs-ent one vear later. After swaiting a sear.ue fed the predictions. together with ourexplanations, back to each site. andasked swhat had happened Iil sesensites. the prediction swas highl! eon-finned; in three it was moderate to high:and in onhs hto sites was the predictiononly moderatelh confirmed. Tlere ssereno reversals or dosviright failures ofprediction.

Tlhis finding reassures us that themodel represents a good uiiderstandiigof institutionalization dsnanmics in these12 sites;. e ulra clcd d the m-istcn toour osin satisfaction. Will the modelapply to other sites as well' Thlat re-mains to be seen. but wse are optilmisticand encourage much wider testing ofthe model in other schools as the! copeu-ith the demarlds of institutionializa-tiol.

Another crucial elemient remainledlatent in our model. It s-as clear in oursites that adminlistrators and teacherslived in separate worlds. Adninistratorspush. demanid support. and thinkabout the organization: teachers react.get involscd. struggle ith the dcmandsof the innovation. and think abouttheir lives wsith studcnts. It s-as crnsclear that an underlying \ariablc v-ccalled teacher-administrator harmonv wascritical for success. VW'orkilng rclationsbetween administrators and teachers hadto be clear and supporti5-e enoughithat the pressures and stresses of incor-porating somethiing necs could bc man-aged together. Thus. both tcacher imas-tenr/comminitmcnt and admiinistratiscaction are critical for institutioializationand linkage betseen thein can beachicecd. 1

'Michael Hluberianl and il\-self. w-ithBe-crl- Lo! Ta!lor and Jo Ann CGoldberg.

-The case reports arc a-ailable at cost fromThe NETWORK, Inc., 290 S. lain St.Andover. MNA 0)1810. The matrsi and net-wsork methods cwere newhlc developed b\ uis.and appear along \s-ith manii other qualita-tive data anlalsis teelhiiques inll Miles alndHuberman (1983. 1984).

Reterences

Bermali. P "Fdlcahtonal Change: sinImplcmcntahoni Paradigm." Inl ImprovrnmSchools: I( sing \' hat We Knowu Idited b R.I.chnrlig and \1 Kiane B'ccrl Hills. Cal-if: Sage. 1981

fBeninai. P . and MIlantghlinl. Feder-al Programs Supporting l ducatronlalChange. \'ol -: Factors .ffecting Implemen-tation and Continuation: Vol h: Imple-menting and Sustaining Innoeatnonn SantaMonica. Calif: Raid Corporatonl. 19-

GCorbet. i. D . D)ausol. I . -. and F'ire-stoie. \V. A. To Each Its Own: SchoolContext and School Change PhiladelpiluResearch for Better Schools. 198'

Crandall. D P.. and associatc s People.Policies and Practices: Ixamining the Chainof School Impromement A.sndmcr. MassThe NET 1VORK. Inc . 19h8

Fnillan. M. The Mleaning of 1 ducationalChange Ns Y'ork:: 1 achers College Prss.alnd Toronto: ()ISE Publications. I9h'

Glascr. E "l)urabilit of hllmOahOlls inHuman Senicc Organizationlss" nonledee3 t 19811: 16--IS,

Howes. N s. "I Contngenc \Iod(C forPredichng Inshtuthonalit ahoo i of mlrns a-hons kcross I)icrgecnt ()rganiahtolns'" l-han\s: State 'lliscrsith of \Ne YoI'rk. 19--

Hubcrman. i M.I. anld Crandall. 1) PPeople. Policies and Practices: IExamining theChain of School Imprormsnent \ l IX- Im-plications for .ction Aindovcr. lass ,:heNET'(ORK. Inc.. 1982

Hubcnialln. M.. and \lilk. \1. BPeople. Policies and Practices: Ixamrning theChain of School Impmrnement \ ol. I In-novation 'p Close: A Field Studs In 12School Settings Anldover. lass lihc \F:1-WO)RK. Inc. 19h'.

Lcake. J: oonrc. \\ .: Pankirat. R:. aidTanner. I. National Follou-Utp \%orshops:Institutionalization. O()aha. \bnr: Clntcrfor Urban Eduncahon. 19-S

louis. K. S.: Roscnblunl. S: Noliihnr.J. A.: arid others Strategies for Knowledge'Use and School Impnroement. \\ ashingtli.

D C: Natiolal Ilstn.,tc of F-dncathn.1981

liles. XI B.. and Ilber.miall. \1.Analhzing Qualitative Data: A Sourrebourof New Methods. Nes York C(enter forPolicA Research. 198. Bcscrlh Hills. CalifSage. 1984.

Rogcrs. E. Diffusion ot Innovations. \es'ork Xlacmillanl. 1962

Yin. R. K.. sith Qulick. S. K.: Baternin.P. M.: and liarks. E 1. Changing U'rbanBureaucracies: How ene Practices BeromeRoutinized. Santa Mlonica. Calif.: RandCorporation. 19-8: and lcxington. Mla.:D. C. Heath. 199.

7tmiran. G.: Duncan. R.: anld H1olhbc. IInnovations and Organizations. Neu \YorkWilec. 1973

NOvLMe�.i 1983 19

NOVEMBER 1983 19

Page 7: Unraveling the Mystery of Institutionalization

Copyright © 1983 by the Association for Supervision and Curriculum Development. All rights reserved.