Download - The History of Psychological Testing

1

1The History ofPsychological Testing

T O P I C 1A The Origins of Psychological TestingThe Importance of TestingCase Exhibit 1.1 The Consequences of Test ResultsRudimentary Forms of Testing in China in 2200 B.C.Psychiatric Antecedents of Psychological TestingThe Brass Instruments Era of TestingChanging Conceptions of Mental Retardation in the 1800sInfluence of Binet’s Early Research upon His TestBinet and Testing for Higher Mental ProcessesThe Revised Scales and the Advent of IQSummary

CH A P T E R

The history of psychological testing is a fasci-nating story and has abundant relevance to

present-day practices. After all, contemporary testsdid not spring from a vacuum; they evolved slowlyfrom a host of precursors introduced over the lastone hundred years. Accordingly, Chapter 1 featuresa review of the historical roots of present-day psy-chological tests. In Topic 1A, The Origins of Psy-chological Testing, we focus largely on the effortsof European psychologists to measure intelligenceduring the late nineteenth century and pre–WorldWar I era. These early intelligence tests and their

successors often exerted powerful effects on theexaminees who took them, so the first topic alsoincorporates a brief digression documenting thepervasive importance of psychological test results.Topic 1B, Early Testing in the United States, cata-logues the profusion of tests developed by Ameri-can psychologists in the first half of the twentiethcentury.

Psychological testing in its modern form origi-nated little more than one hundred years ago in lab-oratory studies of sensory discrimination, motorskills, and reaction time. The British genius Francis

CH01.QXD 6/12/2003 8:50 AM

Galton (1822–1911) invented the first battery oftests, a peculiar assortment of sensory and motormeasures, which we review in the following. TheAmerican psychologist James McKeen Cattell(1860–1944) studied with Galton and then, in 1890,proclaimed the modern testing agenda in his classicpaper entitled “Mental Tests and Measurements.”He was tentative and modest when describing thepurposes and applications of his instruments:

Psychology cannot attain the certainty and exact-ness of the physical sciences, unless it rests on afoundation of experiment and measurement. A stepin this direction could be made by applying a seriesof mental tests and measurements to a large num-ber of individuals. The results would be of consid-erable scientific value in discovering the constancyof mental processes, their interdependence, andtheir variation under different circumstances. Indi-viduals, besides, would find their tests interesting,and, perhaps, useful in regard to training, mode oflife or indication of disease. The scientific andpractical value of such tests would be much in-creased should a uniform system be adopted, sothat determinations made at different times andplaces could be compared and combined. (Cattell,1890)

Cattell’s conjecture that “perhaps” tests wouldbe useful in “training, mode of life or indication ofdisease” must certainly rank as one of the propheticunderstatements of all time. Anyone reared in theWestern world knows that psychological testinghas emerged from its timid beginnings to becomea big business and a cultural institution that per-meates modern society. To cite just one example,consider the number of standardized achievementand ability tests administered in the school systemsof the United States. Although it is difficult to ob-tain exact data on the extent of such testing, an es-timate of 200 million per year is probably notextreme (Medina & Neill, 1990). Of course, thetotal number of tests administered yearly also in-cludes millions of personality tests and untoldnumbers of the thousands of other kinds of testsnow in existence (Conoley & Kramer, 1989, 1992;Mitchell, 1985; Sweetland & Keyser, 1987). Thereis no doubt that testing is pervasive. But does itmake a difference?

THE IMPORTANCE OF TESTING

Tests are used in almost every nation on earth forcounseling, selection, and placement. Testing oc-curs in settings as diverse as schools, civil service,industry, medical clinics, and counseling centers.Most persons have taken dozens of tests andthought nothing of it. Yet, by the time the typical in-dividual reaches retirement age, it is likely that psy-chological test results will help shape his or herdestiny. The deflection of the life course by psy-chological test results might be subtle, such aswhen a prospective mathematician qualifies for anaccelerated calculus course based on tenth-gradeachievement scores. More commonly, psychologi-cal test results alter individual destiny in profoundways. Whether a person is admitted to one collegeand not another, offered one job but refused a sec-ond, diagnosed as depressed or not—all such de-terminations rest, at least in part, on the meaning oftest results as interpreted by persons in authority.Put simply, psychological test results change lives.For this reason it is prudent—indeed, almostmandatory—that students of psychology learnabout the contemporary uses and occasional abusesof testing. In Case Exhibit 1.1, the life-altering af-termath of psychological testing is illustrated bymeans of several true case history examples.

The importance of testing is also evident fromhistorical review. Students of psychology generallyregard historical issues as dull, dry, and pedantic,and sometimes these prejudices are well deserved.After all, many textbooks fail to explain the rele-vance of historical matters and provide only vaguesketches of early developments in mental testing.As a result, students of psychology often concludeincorrectly that historical issues are boring andirrelevant.

In reality, the history of psychological testing isa captivating story that has substantial relevance topresent-day practices. Historical developments arepertinent to contemporary testing for the followingreasons:

1. A review of the origins of psychological testinghelps explain current practices that might other-

2 CHAPTER 1 THE HISTORY OF PSYCHOLOGICAL TESTING

CH01.QXD 6/12/2003 8:50 AM

wise seem arbitrary or even peculiar. For exam-ple, why do many current intelligence tests in-corporate a seemingly nonintellective capacity,namely, short-term memory for digits? The an-swer is, in part, historical inertia—intelligencetests have always included a measure of digitspan.

2. The strengths and limitations of testing also standout better when tests are viewed in historical con-text. The reader will discover, for example, that

modern intelligence tests are exceptionally goodat predicting school failure—precisely becausethis was the original and sole purpose of the firstsuch instrument developed in Paris, France, atthe turn of the twentieth century.

3. Finally, the history of psychological testing con-tains some sad and regrettable episodes thathelp remind us not to be overly zealous in ourmodern-day applications of testing. For exam-ple, based on the misguided and prejudicial

TOPIC 1A THE ORIGINS OF PSYCHOLOGICAL TESTING 3

THE CONSEQUENCES OF TEST RESULTS

The importance of psychological testing is best illustrated by example. Con-sider these brief vignettes:

• A shy, withdrawn 7-year-old girl is administered an IQ test by a school psy-chologist. Her score is phenomenally higher than the teacher expected. Thestudent is admitted to a gifted and talented program where she blossoms intoa self-confident and gregarious scholar.

• Three children in a family living near a lead smelter are exposed to the toxiceffects of lead dust and suffer neurological damage. Based in part on psy-chological test results that demonstrate impaired intelligence and shortenedattention span in the children, the family receives an $8 million settlementfrom the company that owns the smelter.

• A candidate for a position as police officer is administered a personality in-ventory as part of the selection process. The test indicates that the candidatetends to act before thinking and resists supervision from authority figures.Even though he has excellent training and impresses the interviewers, thecandidate does not receive a job offer.

• A student, unsure of what career to pursue, takes a vocational interest in-ventory. The test indicates that she would like the work of a pharmacist. Shesigns up for a prepharmacy curriculum but finds the classes to be both diffi-cult and boring. After three years, she abandons pharmacy for a major in dance, frustrated that she still faces three more years of college to earn adegree.

• An applicant to graduate school in clinical psychology takes the MinnesotaMultiphasic Personality Inventory (MMPI). His recommendations and gradepoint average are superlative, yet he must clear the final hurdle posed by theMMPI. His results are reasonably normal but slightly defensive; by a narrowvote, the admissions committee extends him an invitation. Ironically, this isthe only graduate school to admit him—nineteen others turn him down. Heaccepts the invitation and becomes enchanted with the study of psychologi-cal assessment. Many years later, he writes this book.

CASE EXHIBIT1.1

CH01.QXD 6/12/2003 8:50 AM

application of intelligence test results, severalprominent psychologists helped ensure passageof the Immigration Restriction Act of 1924.

In later chapters, we examine the principles ofpsychological testing, investigate applications inspecific fields (e.g., personality, intelligence,neuropsychology), and reflect on the social andlegal consequences of testing. However, the readerwill find these topics more comprehensible whenviewed in historical context. So, for now, we beginat the beginning by reviewing rudimentary formsof testing that existed over four thousand years agoin imperial China.

RUDIMENTARY FORMS OF TESTINGIN CHINA IN 2200 B .C .

Although the widespread use of psychological test-ing is largely a phenomenon of the twentieth cen-tury, historians note that rudimentary forms oftesting date back to at least 2200 B.C. when the Chi-nese emperor had his officials examined every thirdyear to determine their fitness for office (Bowman,1989; Chaffee, 1985; DuBois, 1970; Franke, 1963;Lai, 1970; Teng, 1942–43). Such testing was modi-fied and refined over the centuries until writtenexams were introduced in the Han dynasty (202B.C.–A.D. 200). Five topics were tested: civil law,military affairs, agriculture, revenue, and geography.

The Chinese examination system took its finalform about 1370 when proficiency in the Confucianclassics was emphasized. In the preliminary exam-ination, candidates were required to spend a dayand a night in a small isolated booth, composing es-says on assigned topics and writing a poem. The 1to 7 percent who passed moved up to the districtexaminations, which required three separate ses-sions of three days and three nights.

The district examinations were obviously gruel-ing and rigorous, but this was not the final level. The1 to 10 percent who passed were allowed the privi-lege of going to Peking for the final round of exam-inations. Perhaps 3 percent of this final group passedand became mandarins, eligible for public office.

Although the Chinese developed the externaltrappings of a comprehensive civil service exami-

nation program, the similarities between their tra-ditions and current testing practices are, in themain, superficial. Not only were their testing prac-tices unnecessarily grueling, the Chinese also failedto validate their selection procedures. Nonetheless,it does appear that the examination program incor-porated relevant selection criteria. For example, inthe written exams beauty of penmanship wasweighted very heavily. Given the highly stylisticfeatures of Chinese written forms, good penman-ship was no doubt essential for clear, exact com-munication. Thus, penmanship was probably arelevant predictor of suitability for civil service em-ployment. In response to widespread discontent,the examination system was abolished by royal de-cree in 1906 (Franke, 1963).

PSYCHIATRIC ANTECEDENTS OF PSYCHOLOGICAL TESTING

Most historians trace the beginnings of psycholog-ical testing to the experimental investigation of in-dividual differences that flourished in Germany andGreat Britain in the late 1800s. There is no doubtthat early experimentalists such as Wilhelm Wundt,Francis Galton, and James McKeen Cattell laid thefoundations for modern-day testing, and we will re-view their contributions in detail. But psychologi-cal testing owes as much to early psychiatry as itdoes to the laboratories of experimental psychol-ogy. In fact, the examination of the mentally illaround the middle of the nineteenth century re-sulted in the development of numerous early tests(Bondy, 1974). These early tests featured the ab-sence of standardization and were consequentlyrelegated to oblivion. They were nonetheless influ-ential in determining the course of psychologicaltesting, so it is important to mention a few typicaldevelopments from this era.

In 1885, the German physician Hubert vonGrashey developed the antecedent of the memorydrum as a means of testing brain-injured patients.His subjects were shown words, symbols, or pic-tures through a slot in a sheet of paper that wasmoving slowly over the stimuli. Grashey found thatmany patients could recognize stimuli in their to-tality but could not identify them when shown


CH01.QXD 6/12/2003 8:50 AM

through the moving slot. Shortly thereafter, theGerman psychiatrist Conrad Rieger developed anexcessively ambitious test battery for brain dam-age. His battery took over 100 hours to administerand soon fell out of favor.

In summary, early psychiatry contributed to themental test movement by showing that standard-ized procedures could help reveal the nature andextent of symptoms in the mentally ill and brain-injured patients. Most of the early tests developedby psychiatrists faded into oblivion, but a few pro-cedures were standardized and perpetuate them-selves in modern variations (Bondy, 1974).

THE BRASS INSTRUMENTS ERA OF TESTING

Experimental psychology flourished in the late1800s in continental Europe and Great Britain. Forthe first time in history, psychologists departedfrom the wholly subjective and introspective meth-ods that had been so fruitlessly pursued in the pre-ceding centuries. Human abilities were insteadtested in laboratories. Researchers used objectiveprocedures that were capable of replication. Gonewere the days when rival laboratories would haveraging arguments about “imageless thought,” onegroup saying it existed, another group saying thatsuch a mental event was impossible.

Even though the new emphasis on objectivemethods and measurable quantities was a vast im-provement over the largely sterile mentalism thatpreceded it, the new experimental psychology wasitself a dead end, at least as far as psychologicaltesting was concerned. The problem was that theearly experimental psychologists mistook simplesensory processes for intelligence. They used as-sorted brass instruments to measure sensory thresh-olds and reaction times, thinking that such abilitieswere at the heart of intelligence. Hence, this periodis sometimes referred to as the Brass Instrumentsera of psychological testing.

In spite of the false start made by early experi-mentalists, at least they provided psychology withan appropriate methodology. Such pioneers asWundt, Galton, Cattell, and Clark Wissler showedthat it was possible to expose the mind to scientific

scrutiny and measurement. This was a fatefulchange in the axiomatic assumptions of psychol-ogy, a change that has stayed with us to the currentday.

Most sources credit Wilhelm Wundt (1832–1920) with founding the first psychological labora-tory in 1879 in Leipzig, Germany. It is less wellrecognized that he was measuring mental processesyears before, at least as early as 1862, when he ex-perimented with his thought meter (Diamond,1980). This device was a calibrated pendulum withneedles sticking off from each side. The pendulumwould swing back and forth, striking bells with theneedles. The observer’s task was to take note of theposition of the pendulum when the bells sounded.Of course, Wundt could adjust the needles before-hand and thereby know the precise position of thependulum when each bell was struck. Wundtthought that the difference between the observedpendulum position and the actual position wouldprovide a means of determining the swiftness ofthought of the observer.

Wundt’s analysis was relevant to a longstandingproblem in astronomy. The problem was that twoor more astronomers simultaneously using thesame telescope (with multiple eyepieces) would re-port different crossing times as the stars movedacross a grid line on the telescope. Even in Wundt’stime, it was a well-known event in the history ofscience that Kinnebrook, an assistant at the RoyalObservatory in England, had been dismissed in1796 because his stellar crossing times were nearlya full second too slow (Boring, 1950). Wundt’sanalysis offered another explanation that did not as-sume incompetence on the part of anyone. Put sim-ply, Wundt believed that the speed of thought mightdiffer from one person to the next:

For each person there must be a certain speed ofthinking, which he can never exceed with his givenmental constitution. But just as one steam enginecan go faster than another, so this speed of thoughtwill probably not be the same in all persons.(Wundt, 1862, as translated in Rieber, 1980)

This analysis of telescope reporting times seemssimplistic by present-day standards and overlooksthe possible contribution of such factors as attention,


CH01.QXD 6/12/2003 8:50 AM

motivation, and self-correcting feedback from priortrials. On the positive side, this was at least an em-pirical analysis that sought to explain individualdifferences instead of trying to explain them away.And that is the relevance to current practices inpsychological testing. However crudely, Wundtmeasured mental processes and begrudgingly ac-knowledged individual differences.1

Galton and the First Battery of Mental Tests

Sir Francis Galton (1822–1911) pioneered the newexperimental psychology in nineteenth-centuryGreat Britain. Galton was obsessed with measure-ment, and his intellectual career seems to have beendominated by a belief that virtually anything wasmeasurable. His attempts to measure intellect bymeans of reaction time and sensory discriminationtasks are well known. Yet, to appreciate his wide-ranging interests, the reader should be apprised thatGalton also devised techniques for measuringbeauty, personality, the boringness of lectures, andthe efficacy of prayer, to name but a few of the en-deavors that his biographer has catalogued in elab-orate detail (Pearson 1914, 1924, 1930ab).

Galton was a genius who was more interestedin the problems of human evolution than in psy-chology per se (Boring, 1950). His two most influ-ential works were Hereditary Genius (1869), anempirical analysis purporting to prove that geneticfactors were overwhelmingly important for the at-tainment of eminence, and Inquiries into HumanFaculty and Its Development (1883), a disparate se-ries of essays that emphasized individual differ-ences in mental faculties.

Boring (1950) regards Inquiries as the begin-ning of the mental test movement and the advent ofthe scientific psychology of individual differences.The book is a curious mixture of empirical researchand speculative essays on topics as diverse as “justperceptible differences” in lifted weight and di-minished fertility among inbred animals. There is,

nonetheless, a common theme uniting these diverseessays; Galton demonstrates time and again that in-dividual differences not only exist but are objec-tively measurable.

Galton borrowed the time-consuming psy-chophysical procedures practiced by Wundt andothers on the European continent and adapted them to a series of simple and quick sensorimotormeasures. Thus, he continued the tradition of brassinstruments mental testing but with an impor-tant difference: his procedures were much moreamenable to the timely collection of data from hun-dreds if not thousands of subjects. Because of hisefforts in devising practicable measures of individ-ual differences, historians of psychological testingusually regard Galton as the father of mental test-ing (Goodenough, 1949; Boring, 1950).

To further his study of individual differences,Galton set up a psychometric laboratory in Londonat the International Health Exhibition in 1884. Itwas later transferred to the London Museum, whereit was maintained for six years. Various anthropo-metric and psychometric measures were arrangedon a long table at one side of a narrow room. Sub-jects were admitted at one end for threepence andgiven successive tests as they moved down the table.At least 17,000 individuals were tested during the1880s and 1890s. About 7,500 of the individual datarecords have survived to the present day (Johnsonet al., 1985).

The tests and measures involved both the phys-ical and behavioral domains. Physical characteris-tics assessed were height, weight, head length, headbreadth, arm span, length of middle finger, andlength of lower arm, among others. The behavioraltests included strength of hand squeeze determinedby dynamometer, vital capacity of the lungs mea-sured by spirometer, visual acuity, highest audibletone, speed of blow, and reaction time (RT) to bothvisual and auditory stimuli.

Ultimately, Galton’s simplistic attempts togauge intellect with measures of reaction time andsensory discrimination proved fruitless. Nonethe-less, he did provide a tremendous impetus to thetesting movement by demonstrating that objectivetests could be devised and that meaningful scorescould be obtained through standardized procedures.


1. This emphasis upon individual differences was rare forWundt. He is more renowned for proposing common laws ofthought for the average adult mind.

CH01.QXD 6/12/2003 8:50 AM

Cattell Imports Brass Instruments to the United States

James McKeen Cattell (1860–1944) studied thenew experimental psychology with both Wundt andGalton before settling at Columbia Universitywhere, for twenty-six years, he was the undisputeddean of American psychology. With Wundt, he dida series of painstakingly elaborate RT studies(1880–1882), measuring with great precision thefractions of a second presumably required for dif-ferent mental reactions. He also noted, almost inpassing, that he and another colleague had smallbut consistent differences in RT. Cattell proposed toWundt that such individual differences ought to bestudied systematically. Although Wundt acknowl-edged individual differences, he was philosophi-cally more inclined to study general features of themind, and he offered no support for Cattell’s pro-posal (Fancher, 1985).

But Cattell received enthusiastic support for hisstudy of individual differences from Galton, whohad just opened his psychometric laboratory in Lon-don. After corresponding with Galton for a fewyears, Cattell arranged for a two-year fellowship atCambridge so that he could continue the study of in-dividual differences. Cattell opened his own researchlaboratory and developed a series of tests that weremainly extensions and additions to Galton’s battery.

Cattell (1890) invented the term mental test in hisfamous paper entitled “Mental Tests and Measure-ments.” This paper described his research program,detailing ten mental tests he proposed for use with thegeneral public. These tests were clearly a reworkingand embellishment of the Galtonian tradition:

Strength of hand squeeze as measured by dynamometer

Rate of hand movement through a distance of50 centimeters

Two-point threshold for touch—minimum dis-tance at which two points are still perceivedas separate

Degree of pressure needed to cause pain—rub-ber tip pressed against the forehead

Weight differentiation—discern the relativeweights of identical-looking boxes varyingby one gram from 100 to 110 grams

Reaction time for sound—using a device simi-lar to Galton’s

Time for naming colorsBisection of a 50-centimeter lineJudgment of 10 seconds of timeNumber of letters repeated on one hearing

Strength of hand squeeze seems a curious addi-tion to a battery of mental tests, a point that Cattell(1890) addressed directly in his paper. He was of theopinion that it was impossible to separate bodily en-ergy from mental energy. Thus, in Cattell’s view, anostensibly physiological measure such as dyna-mometer pressure was an index of one’s mentalpower as well. Clearly, the physiological and sen-sory bias of the entire test battery reflects itsstrongly Galtonian heritage (Fancher, 1985).

In 1891, Cattell accepted a position at Colum-bia University, at that time the largest university inthe United States. His subsequent influence onAmerican psychology was far in excess of his in-dividual scientific output and was expressed inlarge part through his numerous and influential stu-dents (Boring, 1950). Among his many famousdoctoral students and the years of their degreeswere E. L. Thorndike (1898) who made monu-mental contributions to learning theory and educa-tional psychology; R. S. Woodworth (1899) whowas to author the very popular and influential Ex-perimental Psychology (1938); and E. K. Strong(1911) whose Vocational Interest Blank—since re-vised—is still in wide use. But among Cattell’s stu-dents, it was probably Clark Wissler (1901) whohad the greatest influence on the early history ofpsychological testing.

Wissler obtained both mental test scores andacademic grades from more than 300 students atColumbia University and Barnard College. His goalwas to demonstrate that the test results could pre-dict academic performance. With our early twenty-first-century perspective on research and testing, itseems amazing that the early experimentalistswaited so long to do such basic validational re-search. Wissler’s (1901) results showed virtually notendency for the mental test scores to correlate withacademic achievement. For example, class standingcorrelated .16 with memory for number lists, –.08


CH01.QXD 6/12/2003 8:50 AM

with dynamometer strength, .02 with color naming,and –.02 with reaction time. The highest correlation(.16) was statistically significant because of thelarge sample size. However, so humble a correla-tion carries with it very little predictive utility.2

Also damaging to the brass instruments testingmovement was the very modest correlations be-tween the mental tests themselves. For example,color naming and hand movement speed correlatedonly .19, while RT and color naming correlated–.15. Several physical measures such as head size(a holdover measure from the Galton era) were, notsurprisingly, also uncorrelated with the various sen-sory and RT measures.

With the publication of Wissler’s (1901) discouraging results, experimental psychologistslargely abandoned the use of RT and sensory dis-crimination as measures of intelligence. From onestandpoint, this turning away from the brass instru-ments approach was a desirable development in thehistory of psychological testing. The way wasthereby paved for immediate acceptance of AlfredBinet’s more sensible and useful measures ofhigher mental processes.

But in other respects, the abandonment of RTand sensory measures was premature and unfortu-nate. After all, by contemporary standards Wissler’sresearch methods revealed an extraordinary psy-chometric naivete. By using only bright collegestudents as subjects, Wissler had inadvertently in-troduced an extreme restriction of range, whichwould invariably reduce the size of his correlations.If a more heterogeneous sample of subjects hadbeen used, the correlations would have been sub-stantially larger. In addition, certain measures suchas RT were inherently unreliable because of thesmall number of trials per subject. Such unreliabil-ity in a measure also places a severe restriction onthe upper bounds of correlation coefficients.

If Wissler’s (1901) negative findings had beenmore skeptically scrutinized, it might not have beena full 70 years later until RT was resurrected as apotentially useful intellectual measure. Correla-tions of –.40 between complex forms of RT and in-telligence are not at all uncommon (Jensen, 1982).3

But that is getting ahead of the story. The morecommon reaction among psychologists in the early1900s was to begrudgingly conclude that Galtonhad been wrong in attempting to infer complexabilities from simple ones. Goodenough (1949) haslikened Galton’s approach to “inferring the natureof genius from the nature of stupidity or the quali-ties of water from those of the hydrogen andoxygen of which it is composed.” The academicpsychologists apparently agreed with her, andAmerican attempts to develop intelligence tests vir-tually ceased at the turn of the twentieth century.For his own part, Wissler was apparently so dis-couraged by his results that he immediatelyswitched to anthropology, where he became astrong environmentalist in explaining differencesbetween ethnic groups.

The void created by the abandonment of theGaltonian tradition did not last for long. In Europe,Alfred Binet was on the verge of a major break-through in intelligence testing. Binet introduced hisscale of intelligence in 1905, and shortly thereafterH. H. Goddard imported it to the United States,where it was applied in a manner that Gould (1981)has described as “the dismantling of Binet’s inten-tions in America.” Whether early twentieth-centuryAmerican psychologists subverted Binet’s inten-tions is an important question that we review in thenext topic. First, we examine the social changes innineteenth-century Europe that created the neces-sity for practical intelligence tests.

CHANGING CONCEPTIONS OFMENTAL RETARDATION IN THE 1800S

Many great inventions have been developed in re-sponse to the practical needs created by changes in


2. We discuss the correlation coefficient in more detail inTopic 3B, Concepts of Reliability. By way of quick preview, cor-relations can range from –1.0 to +1.0. Values near zero indicatea weak, negligible linear relationship between the two variables.For example, correlations between –.20 and +.20 are generallyof minimal value for purposes of individual prediction. Notealso that negative correlations indicate an inverse relationship.

3. The correlations are negative because low scores on RT areassociated with high scores on intelligence tests.

CH01.QXD 6/12/2003 8:50 AM

societal values. Such is the case with intelligencetests. To be specific, the first such tests were devel-oped by Binet in the early 1900s to help identifychildren in the Paris school system who were un-likely to profit from ordinary instruction. Prior tothis time, there was little interest in the educationalneeds of children with mental retardation. A newhumanism toward those with mental retardation thuscreated the practical problem—identifying thosewith special needs—that Binet’s tests were to solve.

The Western world of the late 1800s was justemerging from centuries of indifference and hos-tility toward the psychiatrically and mentally im-paired. Medical practitioners were just beginningto acknowledge a distinction between individualswith emotional disablities and mental retardation.For centuries, all such social outcasts were givensimilar treatment. In the Middle Ages, they wereoccasionally “diagnosed” as witches and put todeath by burning. Later on, they were alternatelyignored, persecuted, or tortured. In his comprehen-sive history of psychotherapy and psychoanalysis,Bromberg (1959) has an especially graphic chapteron the various forms of maltreatment toward thosewith mental and emotional disabilities, from whichonly one example will be provided here. In 1698, aprominent physician wrote a gruesome book, Fla-gellum Salutis, in which beatings were advocatedas treatment “in melancholia; in frenzy; in paraly-sis; in epilepsy; in facial expression of feeble-minded” (Bromberg, 1959).

By the early 1800s, saner minds began to prevail.Medical practitioners realized that some of those withpsychiatric impairment had reversible illnesses thatdid not necessarily imply diminished intellect,whereas other exceptional persons, those with men-tal retardation, showed a greater developmental con-tinuity and invariably had impaired intellect. Inaddition, a newfound humanism began to influencesocial practices toward individuals with psychologi-cal and mental disabilities. With this humanism therearose a greater interest in the diagnosis and remedia-tion of mental retardation. At the forefront of thesedevelopments were two French physicians, J. E. D.Esquirol and O. E. Seguin, each of whom revolu-tionized thinking about those with mental retarda-

tion, thereby helping to create the necessity forBinet’s tests.

Esquirol and Diagnosis in Mental Retardation

Around the beginning of the nineteenth century,many physicians had begun to perceive the differ-ence between mental retardation (then called id-iocy) and mental illness (often referred to asdementia). J. E. D. Esquirol (1772–1840) was thefirst to formalize the difference in writing. Hisdiagnostic breakthrough was noting that mental re-tardation was a lifelong developmental phenome-non whereas mental illness usually had a moreabrupt onset in adulthood. He thought that mentalretardation was incurable, whereas mental illnessmight show improvement (Esquirol, 1845/1838).

Esquirol placed great emphasis upon languageskills in the diagnosis of mental retardation. This mayoffer a partial explanation as to why Binet’s later testsand the modern-day descendents from them are soheavily loaded on linguistic abilities. After all, theoriginal use of the Binet scales was, in the main, toidentify children with mental retardation who wouldnot likely profit from ordinary schooling.

Esquirol also proposed the first classificationsystem in mental retardation and it should be nosurprise that language skills were the main diag-nostic criteria. He recognized three levels of men-tal retardation: (1) those using short phrases,(2) those using only monosyllables, and (3) thosewith cries only, no speech. Apparently, Esquirol didnot recognize what we would now call mild mentalretardation, instead providing criteria for the equiv-alents of the modern-day classifications of moder-ate, severe, and profound mental retardation.

Seguin and Education of Individuals with Mental Retardation

Perhaps more than any other pioneer in the field ofmental retardation, O. Edouard Seguin (1812–1880) helped establish a new humanism towardthose with mental retardation in the late 1800s. Hehad been a student of Esquirol and had also studiedwith J. M. G. Itard (1774–1838), who is well known


CH01.QXD 6/12/2003 8:50 AM

for his five-year attempt to train the Wild Boy ofAveyron, a feral child who had lived in the woodsfor his first 11 or 12 years (Itard, 1932/1801).

Seguin borrowed from techniques used by Itardand devoted his life to developing educational pro-grams for persons with mental retardation. As earlyas 1838, he had established an experimental classfor such individuals. His treatment efforts earnedhim international acclaim and he eventually cameto the United States to continue his work. In 1866,he published Idiocy, and Its Treatment by the Phys-iological Method, the first major textbook on thetreatment of mental retardation. This book advo-cated a surprisingly modern approach to educationof individuals with mental retardation and eventouched on what would now be called behaviormodification.

Such was the social and historical backgroundthat allowed intelligence tests to flourish. We turnnow to the invention of the modern-day intelligencetest by Alfred Binet. We begin with a discussion ofthe early influences that shaped his famous test.

INFLUENCE OF BINET’S EARLYRESEARCH UPON HIS TEST

As most every student of psychology knows, Al-fred Binet (1857–1911) invented the first modernintelligence test in 1905. What is less well known,but equally important for those who seek an under-standing of his contributions to modern psychol-ogy, is that Binet was a prolific researcher andauthor long before he turned his attentions to intel-ligence testing. The character of his early researchhad a material bearing on the subsequent form ofhis well-known intelligence test. For those whoseek a full understanding of his pathbreaking in-fluence, brief mention of Binet’s early career ismandatory. For more details the reader can consultDuBois (1970), Fancher (1985), Goodenough(1949), Gould (1981), and Wolf (1973).

Binet began his career in medicine, but wasforced to drop out because of a complete emotionalbreakdown. He switched to psychology, where hestudied the two-point threshold and dabbled in the associationist psychology of John Stuart Mill(1806–1873). Later, he selected an apprenticeship

with the neurologist J. M. Charcot (1825–1893) atthe famous Salpetriere Hospital. Thus, for a brieftime Binet’s professional path paralleled that ofSigmund Freud, who also studied hysteria underCharcot. At the Salpetriere Hospital, Binet co-authored (with C. Fere) four studies supposedlydemonstrating that reversing the polarity of a mag-net could induce complete mood changes (e.g.,from happy to sad) or transfer of hysterical paraly-sis (e.g., from left to right side) in a single hypno-tized subject. In response to public criticism fromother psychologists, Binet later published a recan-tation of his findings. This was a painful episodefor Binet, and it sent his career into a temporary de-tour. Nonetheless, he learned two things throughhis embarrassment. First, he never again usedsloppy experimental procedures that allowed forunintentional suggestion to influence his results.Second, he became skeptical of the zeitgeist (spiritof the times) in experimental psychology. Both ofthese lessons were applied when he later developedhis intelligence scales.

In 1891, Binet went to work at the Sorbonne asan unpaid assistant and began a series of studiesand publications that were to define his new “in-dividual psychology” and ultimately to culminatein his intelligence tests. Binet was an ardent exper-imentalist, often using his two daughters to try outexisting and new tests of intelligence. Early on, heflirted with a Cattellian approach to intelligencetesting, using the standard measures of reactiontime and sensory acuity on his two daughters. Theresults were annoyingly inconsistent and difficultto interpret. As might be expected, he found that thereaction times of his children were, on average,much slower than for adults. But on some trials hisdaughters’ performance approached or exceededadult levels. From these findings, Binet concludedthat attention was a key component of intelligence,which was itself a very multifaceted entity. Fur-thermore, he became increasingly disenchantedwith the brass instruments approach to measuringintelligence, which probably explains his subse-quent use of measures of higher mental processes.

In addition, Binet’s sensory-perceptual experi-ments with his children greatly influenced hisviews on proper testing procedures:


CH01.QXD 6/12/2003 8:50 AM

The experimenter is obliged, to a point, to adjusthis method to the subject he is addressing. Thereare certain rules to follow when one experimentson a child, just as there are certain rules for adults,for hysterics, and for the insane. These rules are notwritten down anywhere; each one learns them forhimself and is repaid in great measure. By makingan error and later accounting for the cause, onelearns not to make the mistake a second time. In re-gard to children, it is necessary to be suspicious oftwo principal causes of error: suggestion and fail-ure of attention. This is not the time to speak on thefirst point. As for the second, failure of attention, itis so important that it is always necessary to sus-pect it when one obtains a negative result. Onemust then suspend the experiments and take themup at a more favorable moment, restarting them 10times, 20 times, with great patience. Children, infact, are often little disposed to pay attention to ex-periments which are not entertaining, and it is use-less to hope that one can make them more attentiveby threatening them with punishment. By particu-lar tricks, however, one can sometimes give the ex-periment a certain appeal. (Binet, 1895, quoted inPollack, 1971)

It is interesting to contrast modern-day testingpractices—which go so far as to specify the exactwording the examiner should use—with Binet’s ad-vice to exercise nearly endless patience and use en-tertaining tricks when testing children.

BINET AND TESTING FOR HIGHERMENTAL PROCESSES

In 1896, Binet and his Sorbonne assistant, VictorHenri, published a pivotal review of German andAmerican work on individual differences. In thishistorically important paper, they argued that intel-ligence could be better measured by means of thehigher psychological processes rather than the ele-mentary sensory processes such as reaction time.After several false starts, Binet and Simon eventu-ally settled on the straightforward format of their1905 scales, discussed subsequently.

The character of the 1905 scale owed much toa prior test developed by Dr. Blin (1902) and hispupil, M. Damaye. They had attempted to improvethe diagnosis of mental retardation by using a bat-

tery of assessments in 20 areas such as spoken lan-guage; knowledge of parts of the body; obedienceto simple commands; naming common objects; andability to read, write, and do simple arithmetic.Binet criticized the scale for being too subjective,for having items reflecting formal education, andfor using a yes or no format on many questions(DuBois, 1970). But he was much impressed withthe idea of using a battery of tests, a feature whichhe adopted in his 1905 scales.

In 1904, the Minister of Public Instruction inParis appointed a commission to decide upon theeducational measures that should be undertakenwith those children who could not profit from reg-ular instruction. The commission concluded thatmedical and educational examinations should beused to identify those children who could not learnby the ordinary methods. Furthermore, it was de-termined that these children should be removedfrom their regular classes and given special in-struction suitable to their more limited intellectualprowess. This was the beginning of the special ed-ucation classroom.

It was evident that a means of selecting childrenfor such special placement was needed, and Binetand his colleague Simon were called upon to de-velop a practical tool for just this purpose. Thusarose the first formal scale for assessing the intelli-gence of children.

Goodenough (1949) has outlined the four waysin which the 1905 scale differed from those whichhad been previously constructed.

1. It made no pretense of measuring precisely anysingle faculty. Rather, it was aimed at assessingthe child’s general mental development with aheterogeneous group of tasks. Thus, the aim wasnot measurement, but classification.

2. It was a brief and practical test. The test took lessthan an hour to administer and required little inthe way of equipment.

3. It measured directly what Binet and Simon re-garded as the essential factor of intelligence—practical judgment—rather than wasting timewith lower-level abilities involving sensory,motor, and perceptual elements. They took apragmatic view of intelligence:


CH01.QXD 6/12/2003 8:50 AM

There is in intelligence, it seems to us, a fundamentalagency the lack or alteration of which has the great-est importance for practical life; that is judgement,otherwise known as good sense, practical sense, ini-tiative, or the faculty of adapting oneself. To judgewell, to understand well, to reason well—these arethe essential wellsprings of intelligence. (Binet andSimon, 1905; as translated in Fancher, 1985)

4. The items were arranged by approximate level ofdifficulty instead of content. A rough standard-

ization had been done with 50 normal childrenranging in age from three to 11 years and severalsubnormal and retarded children as well.

The 30 tests on the 1905 scale ranged from ut-terly simple sensory tests to quite complex verbalabstractions. Thus, the scale was appropriate for as-sessing the entire gamut of intelligence—from se-vere mental retardation to high levels of giftedness.The entire scale is outlined in Table 1.1.


TABLE 1.1 The 1905 Binet-Simon Scale

1. Follows a moving object with the eyes.2. Grasps a small object which is touched.3. Grasps a small object which is seen.4. Recognizes the difference between a square of chocolate and a square of wood.5. Finds and eats a square of chocolate wrapped in paper.6. Executes simple commands and imitates simple gestures.7. Points to familiar named objects, e.g., “Show me the cup.”8. Points to objects represented in pictures, e.g., “Put your finger on the window.”9. Names objects in pictures, e.g., “What is this?” [examiner points to a picture of a sign].

10. Compares two lines of markedly unequal length.11. Repeats three spoken digits.12. Compares two weights.13. Shows susceptibility to suggestion.14. Defines common words by function.15. Repeats a sentence of 15 words.16. Tells how two common objects are different, e.g., “paper and cardboard.”17. Names from memory as many as possible of 13 objects displayed on a board for 30 seconds. [This test was later

dropped because it permitted too many possibilities for distraction.]18. Reproduces from memory two designs shown for 10 seconds.19. Repeats a longer series of digits than in item 11 to test immediate memory.20. Tells how two common objects are alike, e.g., “butterfly and flea.”21. Compares two lines of slightly unequal length.22. Compares five blocks to put them in order of weight.23. Indicates which of the previous five weights the examiner has removed. 24. Produces rhymes, e.g., “What rhymes with ‘school’?”25. A word completion test based on those proposed by Ebbinghaus.26. Puts three nouns, e.g., “Paris, river, fortune” (or three verbs) in a sentence.27. Responds to 25 abstract (comprehension) questions, e.g., “When a person has offended you, and comes to offer

his apologies, what should you do?”28. Reverses the hands of a clock.29. After paper folding and cutting, draws the form of the resulting holes.30. Defines abstract words by designating the difference between, e.g., “boredom and weariness.”

Source: Based on translations in Jenkins and Paterson (1961) and Jensen (1980).

CH01.QXD 6/12/2003 8:50 AM

Except for the very simplest tests, which weredesigned for the classification of very low-grade id-iots (an unfortunate diagnostic term that has sincebeen dropped), the tests were heavily weighted to-ward verbal skills, reflecting Binet’s departurefrom the Galtonian tradition.

An interesting point that is often overlooked bycontemporary students of psychology is that Binetand Simon did not offer a precise method for arriv-ing at a total score on their 1905 scale. It is well toremember that their purpose was classification, notmeasurement, and that their motivation was en-tirely humanitarian, namely, to identify those chil-dren who needed special educational placement.By contemporary standards, it is difficult to acceptthe fuzziness inherent in such an approach, but thatmay reflect a modern penchant for quantificationmore than a weakness in the 1905 scale. In fact,their scale was popular among educators in Paris.And, even with the absence of precise quantifica-tion, the approach was successful in selecting can-didates for special classes.

THE REVISED SCALES AND THE ADVENT OF IQ

In 1908, Binet and Simon published a revision ofthe 1905 scale. In the earlier scale, more than halfthe items had been designed for the very retarded,yet the major diagnostic decisions involved olderchildren and those with borderline intellect. To rem-edy this imbalance, most of the very simple itemswere dropped and new items were added at thehigher end of the scale. The 1908 scale had 58 prob-lems or tests, almost double the number from 1905.Several new tests were added, many of which arestill used today: reconstructing scrambled sen-tences, copying a diamond, and executing a se-quence of three commands. Some of the items wereabsurdities that the children had to detect and ex-plain. One such item was amusing to French chil-dren: “The body of an unfortunate girl was found,cut into 18 pieces. It is thought that she killed her-self.” However, this item was very upsetting to someAmerican subjects, demonstrating the importanceof cultural factors in intelligence (Fancher, 1985).

The major innovation of the 1908 scale was theintroduction of the concept of mental level. The testshad been standardized on about 300 normal childrenbetween the ages of 3 and 13 years. This allowedBinet and Simon to order the tests according to theage level at which they were typically passed.Whichever items were passed by 80 to 90 percentof the 3-year-olds were placed in the 3-year level,and similarly on up to age 13. Binet and Simon alsodevised a rough scoring system whereby a basal agewas first determined from the age level at which notmore than one test was failed. For each five tests thatwere passed at levels above the basal, a full year ofmental level was granted. Insofar as partial years ofmental level were not credited and the various agelevels had anywhere from three to eight tests, themethod left much to be desired.

In 1911, a third revision of the Binet-Simonscales appeared. Each age level now had exactlyfive tests. The scale was also extended into theadult range. And with some reluctance, Binet in-troduced new scoring methods that allowed forone-fifth of a year for each subtest passed beyondthe basal level. In his writings, Binet emphasizedstrongly that the child’s exact mental level shouldnot be taken too seriously as an absolute measureof intelligence.

Nonetheless, the idea of deriving a mental levelwas a monumental development that was to influ-ence the character of intelligence testing throughoutthe twentieth century. Within months, what Binetcalled mental level was being translated as mentalage. And testers everywhere, including Binet him-self, were comparing a child’s mental age with thechild’s chronological age. Thus, a 9-year-old whowas functioning at the mental level (or mental age)of a 6-year-old was retarded by three years. Veryshortly, Stern (1912) pointed out that being retardedby three years had different meanings at differentages. A 5-year-old functioning at the 2-year-oldlevel was more impaired than a 13-year-old func-tioning at the 10-year-old level. Stern suggested thatan intelligence quotient computed from the mentalage divided by the chronological age would give abetter measure of the relative functioning of a sub-ject compared to his or her same-aged peers.


CH01.QXD 6/12/2003 8:50 AM

In 1916, Terman and his associates at Stanfordrevised the Binet-Simon scales, producing theStanford-Binet, a successful test that is discussedin a later chapter. Terman suggested multiplying theintelligence quotient by 100 to remove fractions; hewas also the first person to use the abbreviation IQ.Thus was born one of the most popular and con-troversial concepts in the history of psychology.

Binet died in 1911 before the IQ swept Americantesting, so we will never know what he would havethought of this new development based on hisscales. However, Simon, his collaborator, latercalled the concept of IQ a “betrayal” of their scale’soriginal objectives (Fancher, 1985, p. 104), and wecan assume from Binet’s humanistic concern thathe might have held a similar opinion.


1. For better or for worse, psychological testresults possess the power to alter lives. A review ofhistorical trends is crucial if we desire to compre-hend the contemporary influence of psychologicaltests.

2. Rudimentary forms of testing date back to2200 B.C. in China. The Chinese emperors usedgrueling written exams to select officials for civilservice.

3. In the mid- to late 1800s, several physi-cians and psychiatrists developed standardized pro-cedures to reveal the nature and extent of symptomsin the mentally ill and brain-injured. For example,in 1885, Hubert von Grashey developed the pre-cursor to the memory drum to test the visual recog-nition skill of brain-injured patients.

4. Modern psychological testing owes its in-ception to the era of brass instruments psychologythat flourished in Europe during the late 1800s. Bytesting sensory thresholds and reaction times, pio-neer test developers such as Sir Francis Galtondemonstrated that it was possible to measure themind in an objective and replicable manner.

5. Wilhelm Wundt founded the first psycho-logical laboratory in 1879 in Leipzig, Germany. In-cluded among his earlier investigations was his1862 attempt to measure the speed of thought withthe thought meter, a calibrated pendulum with nee-dles sticking off from each side.

6. The first reference to mental tests occurredin 1890 in a classic paper by James McKeen Cat-tell, an American psychologist who had studiedwith Galton. Cattell imported the brass instrumentsapproach to the United States.

7. One of Cattell’s students, Clark Wissler,showed that reaction time and sensory discrimina-tion measures did not correlate with college grades,thereby redirecting the mental-testing movementaway from brass instruments.

8. In the late 1800s, a newfound humanismtoward the mentally retarded, reflected in the diag-nostic and remedial work of French physicians Es-quirol and Seguin, helped create the necessity forearly intelligence tests.

9. Alfred Binet, who was to invent the firsttrue intelligence test, began his career by studyinghysterical paralysis with the French neurologistCharcot. Binet’s claim that magnetism could curehysteria was, to his pained embarrassment, dis-proved. Shortly thereafter, he switched interestsand conducted sensory-perceptual studies, usinghis children as subjects.

10. In 1905, Binet and Simon developed thefirst useful intelligence test in Paris, France. Theirsimple 30-item measure of mainly higher mentalfunctions helped identify schoolchildren who couldnot profit from regular instruction. Curiously, therewas no method for scoring the test.

11. In 1908, Binet and Simon published a re-vised 58-item scale that incorporated the concept ofmental level. In 1911, a third revision of the Binet-Simon scales appeared. Each age level now had ex-actly five tests; the scale extended into the adult range.

12. In 1912, Stern proposed dividing the mental age by the chronological age to obtain anintelligence quotient. In 1916, Terman suggestedmultiplying the intelligence quotient by 100 to re-move fractions. Thus was born the concept of IQ.

SUMMARY

CH01.QXD 6/12/2003 8:50 AM

T O P I C 1B Early Testing in the United StatesEarly Uses and Abuses of Tests in the United StatesThe Invention of Nonverbal Tests in the Early 1900sThe Stanford-Binet: The Early Mainstay of IQGroup Tests and the Classification of WWI Army RecruitsEarly Educational TestingThe Development of Aptitude TestsPersonality and Vocational Testing After WWIThe Origins of Projective TestingThe Development of Interest InventoriesSummary of Major Landmarks in the History of TestingSummary

The Binet-Simon scales helped solve a practi-cal social quandary, namely, how to identify

children who needed special schooling. With thissuccessful application of a mental test, psycholo-gists realized that their inventions could have prag-matic significance for many different segments ofsociety. Almost immediately, psychologists in theUnited States adopted a utilitarian focus. Intelli-gence testing was embraced by many as a reliableand objective response to perceived social prob-lems such as the identification of immigrants withmental retardation and the quick, accurate classifi-cation of Army recruits (Boake, 2002).

Whether these early tests really solved socialdilemmas—or merely exacerbated them—is afiercely debated issue reviewed in the followingsections. One thing is certain: The profusion oftests developed early in the twentieth centuryhelped shape the character of contemporary tests.A review of these historical trends will aid in thecomprehension of the nature of modern tests and abetter appreciation of the social issues raised bythem.

EARLY USES AND ABUSES OF TESTS IN THE UNITED STATES

First Translation of the Binet-Simon Scale

In 1906, Henry H. Goddard was hired by theVineland Training School in New Jersey to doresearch on the classification and education of“feebleminded” children. He soon realized that adiagnostic instrument would be required and wastherefore pleased to read of the 1908 Binet-Simonscale. He quickly set about translating the scale,making minor changes so that it would be applica-ble to American children (Goddard, 1910a).

Goddard (1910b) tested 378 residents of theVineland facility and categorized them by diagno-sis and mental age. He classified 73 residents as id-iots because their mental age was 2 years or lower;205 residents were termed imbeciles with mentalage of 3 to 7 years; and 100 residents were deemedfeebleminded with mental age of 8 to 12 years. It isinstructive to note that originally neutral and descriptive terms for portraying levels of mental retardation—idiot, imbecile, and feebleminded—

15

CH01.QXD 6/12/2003 8:50 AM

have made their way into the everyday lexicon ofpejorative labels. In fact, Goddard made his owncontribution by coining the diagnostic term moron(from the Greek moronia, meaning “foolish”).

Goddard (1911) also tested 1,547 normal chil-dren with his translation of the Binet-Simon scales.He considered children whose mental age was fouror more years behind their chronological age to befeebleminded—these constituted 3 percent of hissample. Considering that all of these children werefound outside of institutions for the retarded, 3 per-cent is rather an alarming rate of mental deficiency.Goddard (1911) was of the opinion that these chil-dren should be segregated so that they would beprevented from “contaminating society.” Theseearly studies piqued Goddard’s curiosity about“feebleminded” citizenry and the societal burdensthey imposed. He also gained a reputation as one ofthe leading experts on the use of intelligence teststo identify persons with impaired intellect. His tal-ents were soon in heavy demand.

The Binet-Simon and Immigration

In 1910, Goddard was invited to Ellis Island by thecommissioner of immigration to help make the ex-amination of immigrants more accurate. A dark andforeboding folklore had grown up around mentaldeficiency and immigration in the early 1900s:

It was believed that the feebleminded were degen-erate beings responsible for many if not most socialproblems; that they reproduced at an alarming rateand menaced the nation’s overall biological fitness;and that their numbers were being incremented byundesirable “new” immigrants from southern andeastern European countries who had largely sup-planted the “old” immigrants from northern andwestern Europe. (Gelb, 1986)

Initially, Goddard was unconcerned about thesupposed threat of feeblemindedness posed by theimmigrants. He wrote that adequate statistics didnot exist and that the prevalent opinions aboutundue percentages of mentally defective immi-grants were “grossly overestimated” (Goddard,1912). However, with repeated visits to Ellis Island,Goddard became convinced that the rates of feeble-

mindedness were much higher than estimated bythe physicians who staffed the immigration service.Within a year, he reversed his opinions entirely andcalled for congressional funding so that Ellis Islandcould be staffed with experts trained in the use ofintelligence tests. In the following decade, Goddardbecame an apostle for the use of intelligence teststo identify feebleminded immigrants. Although hewrote that the rates of mentally deficient immi-grants were “alarming,” he did not join the popularcall for immigration restriction (Gelb, 1986).

The story of Goddard and his concern for the“menace of feeblemindedness,” as Gould (1981)has satirically put it, is often ignored or downplayedin books on psychological testing. The majority oftextbooks on testing do not mention or refer to God-dard at all. The few books that do mention him usu-ally state that Goddard “used the tests in institutionsfor the retarded,” which is surely an understatement.In his influential History of Psychological Testing,DuBois (1970) has a portrait of Goddard but devotesless than one line of text to him.

The fact is that Goddard was one of the most in-fluential American psychologists of the early1900s. Any thoughtful person must therefore won-der why so many contemporary authors have ig-nored or slighted the person who first translated andapplied Binet’s tests in the United States. We willattempt an answer here, based in part on Goddard’soriginal writing, but also relying upon Gould’s(1981) critique of Goddard’s voluminous writingson mental deficiency and intelligence testing. Werefer to Gelb’s (1986) more sympathetic portrayalof Goddard as well.

Perhaps Goddard has been ignored in the text-books because he was a strict hereditarian who con-ceived of intelligence in simple-minded Mendelianterms. No doubt his call for colonization of “mo-rons” so as to restrict their breeding has won himcontemporary disfavor as well. And his insistencethat much undesirable behavior—crime, alco-holism, prostitution—was due to inherited mentaldeficiency also does not sit well with the modernenvironmentalist position.

However, the most likely reason that modernauthors have ignored Goddard is that he exempli-


CH01.QXD 6/12/2003 8:50 AM

fied a large number of early, prominent psycholo-gists who engaged in the blatant misuse of intelli-gence testing. In his efforts to demonstrate that high rates of immigrants with mental retardationwere entering the United States each day, Goddardsent his assistants to Ellis Island to administer hisEnglish translation of the Binet-Simon tests tonewly arrived immigrants. The tests were adminis-tered through a translator, not long after the immi-grants walked ashore. We can guess that many ofthe immigrants were frightened, confused, and dis-oriented. Thus, a test devised in French, then trans-lated to English was, in turn, retranslated back toYiddish, Hungarian, Italian, or Russian; adminis-tered to bewildered farmers and laborers who hadjust endured an Atlantic crossing; and interpretedaccording to the original French norms.

What did Goddard find and what did he make ofhis results? In small samples of immigrants (22 to50), his assistants found 83 percent of the Jews, 80percent of the Hungarians, 79 percent of the Italians,and 87 percent of the Russians to be feebleminded,that is, below age 12 on the Binet-Simon scales(Goddard, 1917). His interpretation of these findingsis, by turns, skeptically cautious and then provoca-tively alarmist. In one place he claims that his study“makes no determination of the actual percentage,even of these groups, who are feebleminded.” Yet,later in the report he states that his figures wouldonly need to be revised by “a relatively smallamount” in order to find the actual percentages offeeblemindedness among immigrant groups. Fur-ther, he concludes that the intelligence of the aver-age immigrant is low, “perhaps of moron grade,” butthen goes on to cite environmental deprivation as theprimary culprit. Simultaneously, Goddard appearsto favor deportation for low IQ immigrants but alsoprovides the humanitarian perspective that we mightbe able to use “moron laborers” if only “we are wiseenough to train them properly.”

There is much, much more to the Goddard eraof early intelligence testing, and the interestedreader is urged to consult Gould (1981) and Gelb(1986). The most important point that we wish tostress here is that—like many other early psychol-ogists—Goddard’s scholarly views were influ-

enced by the social ideologies of his time. Finally,Goddard was a complex scholar who refined andcontradicted his professional opinions on numer-ous occasions. One ironic example: After the dam-age was done and his writings had helped restrictimmigration, Goddard (1928) recanted, concludingthat feeblemindedness was not incurable, and thatthe feebleminded did not need to be segregated ininstitutions.

The Goddard chapter in the history of testingserves as a reminder that even well-meaning per-sons operating within generally accepted socialnorms can misuse psychological tests. We need be ever mindful that disinterested “science” can be harnessed to the goals of a pernicious socialideology.

THE INVENTION OF NONVERBALTESTS IN THE EARLY 1900S

Because of the heavy emphasis of the Binet-Simonscales upon verbal skills, many psychologists real-ized that this new measuring device was not entirelyappropriate for non-English-speaking subjects,illiterates, and those with speech and hearingimpairments. A spate of performance scales there-fore arose in the decade following Goddard’s 1908 translation of the Binet-Simon. Only a briefchronology of nonverbal tests will be supplied here.The interested reader should consult DuBois(1970). In this listing of early performance tests, thereader will surely recognize many instruments andsubtests that are still used today.

The earliest of the performance measures wasthe Seguin form board, an upright stand with de-pressions into which ten blocks of varying shapescould be fitted. This had been used by Seguin as atraining device for individuals with mental retarda-tion, but was subsequently developed as a test byGoddard, and then standardized by R. H. Sylvester(1913). This identical board is still used, with thesubject blindfolded, in the Halstead-Reitan neu-ropsychological test battery (Reitan & Wolfson,1985).

Knox (1914) devised several performance testsfor use with Ellis Island immigrants. His tests

TOPIC 1B EARLY TESTING IN THE UNITED STATES 17

CH01.QXD 6/12/2003 8:50 AM

required absolutely no verbal responses from sub-jects. The examiner demonstrated each task non-verbally to ensure that the subjects understood theinstructions. Included in his tests were a simplewooden puzzle (which Knox referred to as the“moron” test) and the same digit-symbol substitu-tion test which is now found on most of the Wech-sler scales of intelligence.

Several other early performance tests areworthy of brief mention because they have sur-vived to the present day in revised form. Pintnerand Paterson (1917) invented a 15-part scale ofperformance tests that used several form boards,puzzles, and object assembly tests. The object as-sembly test—reassembling cut-up cardboard ver-sions of common objects such as a horse—is amainstay of several contemporary intelligencetests. The Kohs Block Design test (Kohs, 1920),which required the subject to assemble paintedblocks to resemble a pattern, is well known to anymodern tester who uses the Wechsler scales. ThePorteus Maze Test (Porteus, 1915) is a graded se-ries of mazes for which the subject must avoiddead ends while tracing a path from beginning toend. This is a fine instrument that is still availabletoday, but underused.

THE STANFORD-BINET: THE EARLY MAINSTAY OF IQ

While it was Goddard who first translated the Binetscales in the United States, it was Stanford pro-fessor Lewis M. Terman (1857–1956) who popu-larized IQ testing with his revision of the Binetscales in 1916. The new Stanford-Binet, as it wascalled, was a substantial revision, not just an ex-tension, of the earlier Binet scales. Among themany changes that led to the unquestioned prestigeof the Stanford-Binet was the use of the now fa-miliar IQ for expressing test results. The number ofitems was increased to 90, and the new scale wassuitable for those with mental retardation, children,and both normal and “superior” adults. In addition,the Stanford-Binet had clear and well-organized in-structions for administration and scoring. Great

care had been taken in securing a representativesample of subjects for use in the standardization ofthe test. As Goodenough (1949) notes: “The publi-cation of the Stanford Revision marked the end ofthe initial period of experimentation and uncer-tainty. Once and for all, intelligence testing hadbeen put on a firm basis.”

The Stanford-Binet was the standard of intelli-gence testing for decades. New tests were alwaysvalidated in terms of their correlations with thismeasure. It continued its preeminence through re-visions in 1937, and 1960, by which time the Wech-sler scales (Wechsler, 1949, 1955) had begun tocompete with it. The latest revision of the Stanford-Binet was completed in 2003. This test and theWechsler scales are discussed in detail in a laterchapter. It is worth mentioning here that the Wech-sler scales became a quite popular alternative to theStanford-Binet mainly because they provided morethan just an IQ score. In addition to Full Scale IQ,the Wechsler scales provided ten to twelve subtestscores, and a Verbal and Performance IQ. By con-trast, the earlier versions of the Stanford-Binet sup-plied only a single overall summary score, theglobal IQ.

GROUP TESTS AND THECLASSIFICATION OF WWI ARMY RECRUITS

Given the American penchant for efficiency, it wasonly natural that researchers would seek group men-tal tests to supplement the relatively time-consum-ing individual intelligence tests imported fromFrance. Among the first to develop group tests wasPyle (1913), who published schoolchildren normsfor a battery consisting of such well-worn measuresas memory span, digit-symbol substitution, and oralword association (quickly writing down words in re-sponse to a stimulus word). Pintner (1917) revisedand expanded Pyle’s battery, adding to it a timedcancellation test in which the child crossed out theletter a wherever it appeared in a body of text.

But group tests were slow to catch on, partlybecause the early versions still had to be scored


CH01.QXD 6/12/2003 8:50 AM

laboriously by hand. The idea of a completely ob-jective test with a simple scoring key was inconsis-tent with tests such as logical memory for whichthe judgment of the examiner was required in scor-ing. Most amazing of all—at least to anyone whohas spent any time as a student in Americanschools—the multiple-choice question was not yetin general use.

The slow pace of developments in group test-ing picked up dramatically as the United States en-tered World War I in 1917. It was then that RobertM. Yerkes, a well-known psychology professor atHarvard, convinced the U.S. government and theArmy that all of its 1.75 million recruits should begiven intelligence tests for purposes of classifica-tion and assignment (Yerkes, 1919). Immediatelyupon being commissioned into the Army as acolonel, Yerkes assembled a Committee on the Ex-amination of Recruits, which met at the Vinelandschool in New Jersey to develop the new group testsfor the assessment of Army recruits. Yerkes chairedthe committee; other famous members includedGoddard and Terman.

Two group tests emerged from this collabora-tion: the Army Alpha and the Army Beta. It wouldbe difficult to overestimate the influence of theAlpha and Beta upon subsequent intelligence tests.The format and content of these tests inspired de-velopments in group and individual testing fordecades to come. We discuss these tests in some de-tail so that the reader can appreciate their influenceon modern intelligence tests.

The Army Alpha and Beta Examinations

The Alpha was based on the then unpublished workof Otis (1918) and consisted of eight verbally loadedtests for average and high-functioning recruits. Theeight tests were: (1) following oral directions,(2) arithmetical reasoning, (3) practical judgment,(4) synonym–antonym pairs, (5) disarranged sen-tences, (6) number series completion, (7) analogies,and (8) information. Figure 1.1 lists some typicalitems from the Army Alpha examination.

The Army Beta was a nonverbal group test de-signed for use with illiterates and recruits whose

first language was not English. It consisted of var-ious visual-perceptual and motor tests such astracing a path through mazes and visualizing thecorrect number of blocks depicted in a three-dimensional drawing. Figure 1.2 depicts the black-board demonstrations for all eight parts of the Betaexamination.

In order to accommodate illiterate subjects andrecent immigrants who did not comprehend Eng-lish, Yerkes instructed the examiners to use largelypictorial and gestural methods for explaining thetests to the prospective Army recruits. The exam-iner and an assistant stood atop a platform at thefront of the class and engaged in pantomime to ex-plain each of the eight tests. We reproduce here theexact instructions for one test so that the reader canappraise the likely effects of the testing proceduresupon Beta results. Keep in mind that many recruitscould not see or hear the examiner well, and thatsome had never taken a test before. Here is how theexaminers introduced test 6, picture completion, toeach new roomful of potential recruits:

“This is test 6 here. Look. A lot of pictures.” Aftereveryone has found the place, “Now watch.” Exam-iner points to hand and says to demonstrator, “Fixit.” Demonstrator does nothing, but looks puzzled.Examiner points to the picture of the hand, andthen to the place where the finger is missing andsays to demonstrator, “Fix it; fix it.” Demonstratorthen draws in finger. Examiner says “That’s right.”Examiner then points to fish and place for eye andsays, “Fix it.” After demonstrator has drawn miss-ing eye, examiner points to each of the four re-maining drawings and says, “Fix them all.”Demonstrator works samples out slowly and withapparent effort. When the samples are finished ex-aminer says, “All right. Go head. Hurry up!” Dur-ing the course of this test the orderlies walk aroundthe room and locate individuals who are doingnothing, point to their pages and say, “Fix it. Fixthem,” trying to set everyone working. At the endof 3 minutes examiner says, “Stop! But don’t turnover the page.” (Yerkes, 1921)

The Army testing was intended to help segre-gate and eliminate the mentally incompetent, toclassify men according to their mental ability, andto assist in the placement of competent men in


CH01.QXD 6/12/2003 8:50 AM

responsible positions (Yerkes, 1921). However, itis not really clear whether the Army made muchuse of the masses of data supplied by Yerkes andhis eager assistants. A careful reading of hismemoirs reveals that Yerkes did little more thanproduce favorable testimonials from high-rank-

ing officers. In the main, his memoirs say that theArmy could have saved millions of dollars and in-creased its efficiency, if the testing data had beenused.

To some extent, the mountains of test data hadlittle practical impact on the efficiency of the Army


Note: Examinees received verbal instructions for each subtest.

FIGURE 1.1 Sample Items from the Army Alpha ExaminationSource: Reprinted from Yerkes, R. M. (Ed.). (1921). Psychological examining in the United States Army. Memoirs of the National Academy of Sciences, Volume 15. With permission from the National Academy of Sciences, Washington, DC.

FOLLOWING ORAL DIRECTIONS

Mark a cross in the first and also the third circle:� � � � �

ARITHMETICAL REASONING

Solve each problem:How many men are 5 men and 10 men? Answer ( )If 3 1/2 tons of coal cost $21, what will 5 1/2 tons cost? Answer ( )

PRACTICAL JUDGMENT

Why are high mountains covered with snow? Because� they are near the clouds� the sun shines seldom on them� the air is cold there

SYNONYM–ANTONYM PAIRS

Are these words the same or opposite?largess—donation same? or opposite?accumulate—dissipate same? or opposite?

DISARRANGED SENTENCES

Can these words be rearranged to form a sentence?envy bad malice traits are and true? or false?

NUMBER SERIES COMPLETION

Complete the series: 3 6 8 16 18 36 . . . . . .

ANALOGIES

Which choice completes the analogy?tears—sorrow :: laughter— joy smile girls gringranary—wheat :: library— desk books paper librarian

INFORMATION

Choose the best alternative:The pancreas is in the abdomen head shoulder neckThe Battle of Gettysburg was fought in 1863 1813 1778 1812

CH01.QXD 6/12/2003 8:50 AM

because of the resistance of the military mind toscientific innovation. However, it is also true thatthe Army brass had good reason to doubt the va-

lidity of the test results. For example, an internalmemorandum described the use of pantomime inthe instructions to the nonverbal Beta examination:


FIGURE 1.2The Blackboard Demonstrations for All Eight Parts of the Beta ExaminationSource: Reprinted from Yerkes, R. M. (Ed.). (1921).Psychological examining in the United States Army. Memoirs of the National Academy of Sciences, Vol-ume 15. With permission from the National Acad-emy of Sciences, Washington, DC.

CH01.QXD 6/12/2003 8:50 AM

For the sake of making results from the variouscamps comparable, the examiners were ordered tofollow a certain detailed and specific series of bal-let antics, which had not only the merit of beingperfectly incomprehensible and unrelated to mentaltesting, but also lent a highly confusing and dis-tracting mystical atmosphere to the whole perfor-mance, effectually preventing all approach to theattitude in which a subject should be while havinghis soul tested. (cited in Samelson, 1977)

In addition, the testing conditions left much to bedesired, with wave upon wave of recruits usheredin one door, tested, and virtually shoved out theother side. Tens of thousands of recruits received aliteral zero for many subtests, not because theywere retarded but because they couldn’t fathom theinstructions to these enigmatic new instruments.Many recruits fell asleep while the testers gave es-oteric and mysterious pantomime instructions.

On the positive side, the Army testing providedpsychologists with a tremendous amount of expe-rience in the psychometrics of test construction.Thousands of correlation coefficients were com-puted, including the prominent use of multiplecorrelations in the analysis of test data. Test con-struction graduated from an art to a science in a fewshort years.

The Army Tests and Ethnic Differences

Unfortunately, the Army test results were some-times used to substantiate prejudices about variousracial and ethnic groups rather than to dispassion-ately investigate the causes of group differences.For example, in his influential book A Study ofAmerican Intelligence, Brigham (1923) undertooka massive analysis of Alpha and Beta scores forNordic, Mediterranean, and Alpine immigrants.The text is stuffed with ostensibly objective tablesand charts comparing racial and ethnic groups. Forexample, one curious figure in his book depicts theproportion of each immigration sample at or belowthe average of the African American draft. Brighamconcluded that African Americans, Mediterraneanimmigrants, and Alpine immigrants were intellec-tually inferior. He sounded a dire warning that

racial intermixture would inevitably cause a deteri-oration of American intelligence. For example, thecaption to one graph reads, in part:

The distributions of the intelligence scores of theentire Nordic group, the combined Mediterraneanand Alpine groups, and the negro draft. Theprocess of racial intermixture cannot result in any-thing but an average of these elements, with the re-sulting deterioration of American intelligence.(Brigham, 1923)

Seven years later, Brigham (1930) forthrightlydisavowed his earlier views. He cited cultural andlanguage differences as the likely cause of ethnicand racial disparities on the Army tests. He assertedthat comparative studies of national and racialgroups could not be made with existing tests andconcluded that his earlier findings were “withoutfoundation” (Brigham, 1930).

EARLY EDUCATIONAL TESTING

For good or for ill, Yerkes’s grand scheme for test-ing Army recruits helped to usher in the era ofgroup tests. After WWI, inquiries rushed in fromindustry, public schools, and colleges about the po-tential applications of these straightforward teststhat almost anyone could administer and score(Yerkes, 1921). The psychologists who had workedwith Yerkes soon left the service and carried withthem to industry and education their newfound no-tion of paper-and-pencil tests of intelligence.

The Army Alpha and Beta were also releasedfor general use. These tests quickly became theprototypes for a large family of group tests andinfluenced the character of intelligence tests, col-lege entrance examinations, scholastic achievementtests, and aptitude tests. To cite just one specificconsequence of the Army testing, the NationalResearch Council, a government organization ofscientists, devised the National Intelligence Test,which was eventually given to 7 million children in the United States during the 1920s. Thus, suchwell-known tests as the Wechsler scales, the Scho-lastic Aptitude Tests, and the Graduate RecordExam actually have roots that reach back to Yerkes,


CH01.QXD 6/12/2003 8:50 AM

Otis, and the mass testing of Army recruits duringWWI.

The College Entrance Examination Board(CEEB) was established at the turn of the twentiethcentury to help avoid duplication in the testing ofapplicants to U.S. colleges. The early exams hadbeen of the short answer essay format, but this wasto change quickly when C. C. Brigham, a discipleof Yerkes, became CEEB secretary after WWI. In1925, the College Board decided to construct ascholastic aptitude test for use in college admis-sions (Goslin, 1963). The new tests reflected thenow familiar objective format of unscrambling sen-tences, completing analogies, and filling in the nextnumber in a sequence. Machine scoring was intro-duced in the 1930s, making objective group testseven more efficient than before. These tests thenevolved into the present College Board tests, in par-ticular, the Scholastic Aptitude Tests, now knownas the Scholastic Assessment Tests.

The functions of the CEEB were later sub-sumed under the nonprofit Educational Testing Ser-vice (ETS). The ETS directed the development,standardization, and validation of such well-knowntests as the Graduate Record Examination, the Law School Admissions Test, and the Peace CorpsEntrance Tests.

Meanwhile, Terman and his associates at Stan-ford were busy developing standardized achieve-ment tests. The Stanford Achievement Test (SAchT)was first published in 1923; a modern version of itis still in wide use today. From the very beginning,the SAchT incorporated such modern psychomet-ric principles as norming the subtests so thatwithin-subject variability could be assessed and se-lecting a very large and representative standardiza-tion sample.

THE DEVELOPMENT OF APTITUDE TESTS

Aptitude tests measure more specific and delimitedabilities than intelligence tests. Traditionally, intel-ligence tests assess a more global construct such asgeneral intelligence, although there are exceptionsto this trend that will be discussed later. By con-

trast, a single aptitude test will measure just oneability domain, and a multiple aptitude test batterywill provide scores in several distinctive abilityareas.

The development of aptitude tests lagged be-hind that of intelligence tests for two reasons, onestatistical, the other social. The statistical problemwas that a new technique, factor analysis, wasoften needed to discern which aptitudes were pri-mary and therefore distinct from each other. Re-search on this question had been started quite earlyby Spearman (1904) but was not refined until the1930s (Spearman, 1927; Kelley, 1928; Thurstone,1938). This new family of techniques, factoranalysis, allowed Thurstone to conclude that therewere specific factors of primary mental abilitysuch as verbal comprehension, word fluency, num-ber facility, spatial ability, associative memory,perceptual speed, and general reasoning (Thur-stone, 1938; Thurstone & Thurstone, 1941). Morewill be said about this in the later chapters on in-telligence and ability testing. The important pointhere is that Thurstone and his followers thoughtthat global measures of intelligence did not, so tospeak, “cut nature at its joints.” As a result, it wasfelt that such measures as the Stanford-Binet werenot as useful as multiple aptitude test batteries indetermining a person’s intellectual strengths andweaknesses.

The second reason for the slow growth of apti-tude batteries was the absence of a practical appli-cation for such refined instruments. It was not untilWWII that a pressing need arose to select candi-dates who were highly qualified for very difficultand specialized tasks. The job requirements ofpilots, flight engineers, and navigators were veryspecific and demanding. A general estimate of in-tellectual ability, such as provided by the groupintelligence tests used in WWI, was not sufficientto choose good candidates for flight school. Thearmed forces solved this problem by developing aspecialized aptitude battery of 20 tests that wasadministered to men who passed preliminaryscreening tests. These measures proved invaluablein selecting pilots, navigators, and bombadiers, asreflected in the much lower washout rates of men


CH01.QXD 6/12/2003 8:50 AM

selected by test battery instead of the old methods(Goslin, 1963). Such tests are still used widely inthe armed services.

PERSONALITY AND VOCATIONALTESTING AFTER WWI

While such rudimentary assessment methods asthe free association technique had been used be-fore the turn of the twentieth century by Galton,Kraepelin, and others, it was not until WWI thatpersonality tests emerged in a form resemblingtheir contemporary appearance. As has happenedso often in the history of testing, it was once againa practical need that served as the impetus for thisnew development. Modern personality testingbegan when Woodworth attempted to develop aninstrument for detecting Army recruits who weresusceptible to psychoneurosis. Virtually all themodern personality inventories, schedules, andquestionnaires owe a debt to Woodworth’s Per-sonal Data Sheet (1919).

The Personal Data Sheet consisted of 116 ques-tions that the subject was to answer by underliningYes or No. The questions were exclusively of the“face obvious” variety and, for the most part, in-volved fairly serious symptomatology. Representa-tive items included:

• Do ideas run through your head so that you can-not sleep?

• Were you considered a bad boy?• Are you bothered by a feeling that things are not

real?• Do you have a strong desire to commit suicide?

Readers familiar with the Minnesota MultiphasicPersonality Inventory (MMPI) must surely recog-nize the debt that this more recent inventory has toWoodworth’s instrument.

From his account of how the Personal DataSheet was developed (Woodworth, 1951), it is clearthat Woodworth took great care in the selection ofitems. In other respects, though, this instrumentembodies a large dose of psychometric credulity.The most serious problem is simply that a disturbedsubject motivated to look good could do so without

detection; likewise, a normal subject with a fakebad mentality might be categorized as unfit for ser-vice. Modern instruments such as the MMPI haveincorporated various validity scales for detectingsuch response tendencies. The Personal Data Sheet,by contrast, was predicated on the assumption thatsubjects would be honest when responding to thequestions.

The next major development was an inventoryof neurosis, the Thurstone Personality Schedule(Thurstone & Thurstone, 1930). After first cullinghundreds of items answerable in the yes-no-? man-ner from Woodworth’s inventory and other sources,Thurstone rationally keyed items in terms of howthe neurotic would typically answer them. Reflect-ing Thurstone’s penchant for statistical finesse, thisinventory was one of the first to use the method ofinternal consistency whereby each prospective itemwas correlated with the total score on the tenta-tively identified scale to determine whether it be-longed on the scale.

From the Thurstone test sprang the BernreuterPersonality Inventory (Bernreuter, 1931). It was alittle more refined than its Thurstone predecessor,measuring four personality dimensions: neurotictendency, self-sufficiency, introversion-extrover-sion, and dominance-submission. A major innova-tion in test construction was that a single test itemcould contribute to more than one scale.

The Allport-Vernon Study of Values was alsopublished in 1931 (Allport & Vernon, 1931). Thistest was quite different from the others in that itmeasured values instead of psychopathology. Fur-thermore, it adopted a new scoring method, the ip-sative approach, in which the respondent wascompared only with himself or herself regardingthe balance of importance given to six basic values:theoretical, economic, aesthetic, social, political,and religious. The test was devised in such a man-ner that subjects were required to make choices be-tween the six values in specific situations. As aconsequence, the average on the six scales was al-ways the same for each subject. A weakness in onevalue was compensated for by a strength in someother value. Thus, only the relative peaks and val-leys were of interest.


CH01.QXD 6/12/2003 8:50 AM

Any chronology of self-report inventoriesmust surely include the Minnesota MultiphasicPersonality Inventory, or MMPI (Hathaway &McKinley, 1940). This test and its revision, theMMPI-2, are discussed in detail later. It will suf-fice for now to point out that the scales of theMMPI were constructed by the method that Wood-worth pioneered, contrasting the responses of nor-mal and psychiatrically disturbed subjects. Inaddition, the MMPI introduced the use of validityscales to determine fake bad, fake good, and ran-dom response patterns.

THE ORIGINS OF PROJECTIVE TESTING

The projective approach originated with the wordassociation method pioneered by Francis Galton inthe late 1800s. Galton gave himself four seconds tocome up with as many associations as possible to astimulus word, and then categorized his associa-tions as parrotlike, image-mediated, or histrionicrepresentations. This latter category convinced himthat mental operations “sunk wholly below thelevel of consciousness” were at play. Some histori-ans have even speculated that Freud’s applicationof free association as a therapeutic tool in psycho-analysis sprang from Galton’s paper published inBrain in 1879 (Forrest, 1974).

Galton’s work was continued in Germany byWundt and Kraepelin, and finally brought tofruition by Jung (1910). Jung’s test consisted of100 stimulus words. For each word, the subjectwas to reply as quickly as possible with the firstword coming to mind. Kent and Rosanoff (1910)gave the association method a distinctively Amer-ican flavor by tabulating the reactions of 1,000normal subjects to a list of 100 stimulus words.These tables were designed to provide a basis forcomparing the reactions of normal and “insane”subjects.

While the Americans were pursuing the empir-ical approach to objective personality testing, ayoung Swiss psychiatrist, Hermann Rorschach(1884–1922), was developing a completely differ-ent vehicle for studying personality. Rorschach was

strongly influenced by Jungian and psychoanalyticthinking, so it was natural that his new approach fo-cused on the tendency of patients to reveal their in-nermost conflicts unconsciously when respondingto ambiguous stimuli. The Rorschach and otherprojective tests discussed subsequently were pred-icated upon the projective hypothesis: When re-sponding to ambiguous or unstructured stimuli, weinadvertently disclose our innermost needs, fan-tasies, and conflicts.

Rorschach was convinced that people revealedimportant personality dimensions in their responsesto inkblots. He spent years developing just the rightset of ten inkblots and systematically analyzed theresponses of personal friends and different patientgroups (Rorschach, 1921). Unfortunately, he diedonly a year after his monograph was published, andit was up to others to complete his work. Develop-ments in the Rorschach are reviewed later in thetext.

While Rorschach’s test was originally devel-oped to reveal the innermost workings of the ab-normal subject, the TAT, or Thematic ApperceptionTest (Morgan & Murray, 1935), was developed asan instrument to study normal personality. Ofcourse, both have since been expanded for testingwith the entire continuum of human behavior.

The TAT consists of a series of pictures thatlargely depict one or more persons engaged in anambiguous interaction. The subject is shown onepicture at a time and told to make up a story aboutit. He or she is instructed to be as dramatic as pos-sible, to discuss thoughts and feelings, and to de-scribe the past, present, and future of what isdepicted in the picture.

Murray (1938) believed that underlying per-sonality needs, such as the need for achievement,would be revealed by the contents of the stories.Although numerous scoring systems were devel-oped, clinicians in the main have relied upon animpressionistic analysis to make sense of TAT pro-tocols. Modern applications of the TAT are dis-cussed in a later chapter.

The sentence completion technique was alsobegun during this era with the work of Payne (1928). There have been numerous extensions and


CH01.QXD 6/12/2003 8:50 AM

variations on the technique, which consists of givingsubjects a stem such as “I am bored when ———,”and asking them to complete the sentence. Somemodern applications are discussed later, but it canbe mentioned now that the problem of scoring andinterpretation, which vexed early sentence comple-tion test developers, is still with us today.

An entirely new approach to projective testingwas taken by Goodenough (1926), who tried todetermine not just intellectual level, but also theinterests and personality traits of children by ana-lyzing their drawings. Buck’s (1948) test, theHouse-Tree-Person, was a little more standardizedand structured and required the subject to draw ahouse, a tree, and a person. Machover’s (1949) Per-sonality Projection in the Drawing of the HumanFigure was the logical extension of the earlierwork. Figure drawing as a projective approach tounderstanding personality is still used today, and alater chapter discusses modern developments inthis practice.

Meanwhile, projective testing in Europe wasdominated by the Szondi Test, a wacky instrumentbased on wholly faulty premises. Lipot Szondi wasa Hungarian-born Swiss psychiatrist who believedthat major psychiatric disorders were caused byrecessive genes. His test consisted of 48 photo-graphs of psychiatric patients divided into six setsof the following eight types: homosexual, epilep-tic, sadistic, hysteric, catatonic, paranoiac, manic,and depressive (Deri, 1949). From each set of eightpictures, the subject was instructed to select the twopictures he or she liked best and the two dislikedmost. A person who consistently preferred one kindof picture in the six sets was presumed to havesome recessive genes that made him or her havesympathy for the pictured person. Thus, projectivepreferences were presumed to reveal recessivegenes predisposing the individual to specific psy-chiatric disturbances.

Deri (1949) imported the test to the UnitedStates and changed the rationale. She did not arguefor a recessive genetic explanation of picture choicebut explained such preferences on the basis of un-conscious identification with the characteristics of

the photographed patients. This was a more palat-able theoretical basis for the test than the dubiousgenetic theories of Szondi. Nonetheless, empiricalresearch cast doubt on the validity of the SzondiTest, and it shortly faded into oblivion (Borstel-mann & Klopfer, 1953).

THE DEVELOPMENT OF INTEREST INVENTORIES

While the clinicians were developing measures foranalyzing personality and unconscious conflicts,other psychologists were devising measures forguidance and counseling of the masses of morenormal persons. Chief among such measures wasthe interest inventory, which has roots going backto Thorndike’s (1912) study of developmentaltrends in the interests of 100 college students. In1919–1920, Yoakum developed a pool of 1,000items relating to interests from childhood throughearly maturity (DuBois, 1970). Many of these itemswere incorporated in the Carnegie Interest Inven-tory. Cowdery (1926–27) improved and refinedprevious work on the Carnegie instrument by in-creasing the number of items, comparing responsesof three criterion groups (doctors, engineers, andlawyers) with control groups of nonprofessionals,and developing a weighting formula for items. Hewas also the first psychometrician to realize the im-portance of cross validation. He tested his newscales on additional groups of doctors, engineers,and lawyers to ensure that the discriminationsfound in the original studies were reliable groupdifferences rather than capitalizations on errorvariance.

Edward K. Strong (1884–1963) revisedCowdery’s test and devoted 36 years to the devel-opment of empirical keys for the modified instru-ment known as the Strong Vocational InterestBlank (SVIB). Persons taking the test could bescored on separate keys for several dozen occupa-tions, providing a series of scores of immeasurablevalue in vocational guidance. The SVIB becameone of the most widely used tests of all time(Strong, 1927). Its modern version, the Strong In-


CH01.QXD 6/12/2003 8:50 AM

terest Inventory, is still widely used by guidancecounselors.

For decades the only serious competitor to theSVIB was the Kuder Preference Record (Kuder,1934). The Kuder differed from the Strong by forc-ing choices within triads of items. The Kuder wasan ipsative test; that is, it compared the relativestrength of interests within the individual, ratherthan comparing his or her responses to various pro-fessional groups. More recent revisions of theKuder Preference Record include the Kuder Gen-eral Interest Survey and the Kuder Occupational In-

terest Survey (Kuder, 1966; Kuder & Diamond,1979; Zytowski, 1985).

SUMMARY OF MAJOR LANDMARKSIN THE HISTORY OF TESTING

We conclude our historical survey of psychologi-cal testing with a brief tabular summary of land-mark events up to 1950 (Table 1.2). The interestedreader can find a more detailed listing—includinga chronology of post-1950 developments—in Appendix A.


TABLE 1.2 A Summary of Early Landmarks in the History of Testing

2200 B.C.A.D.1862

1884

1890

1901

19051914

1916

1917

1917

19201921

19271939

19421949

Chinese begin civil service examinations.Wilhelm Wundt uses a calibrated pendulum to measure the “speed ofthought.”Francis Galton administers the first test battery to thousands of citizens atthe International Health Exhibit.James McKeen Cattell uses the term mental test in announcing the agendafor his Galtonian test battery.Clark Wissler discovers that Cattellian “brass instruments” tests have nocorrelation with college grades.Binet and Simon invent the first modern intelligence test.Stern introduces the IQ, or intelligence quotient: the mental age divided bychronological age.Lewis Terman revises the Binet-Simon scales, publishes the Stanford-Binet. Revisions appear in 1937, 1960, and 1986.Robert Yerkes spearheads the development of the Army Alpha and Betaexaminations used for testing WWI recruits.Robert Woodworth develops the Personal Data Sheet, the first personalitytest.Rorschach Inkblot test published.Psychological Corporation—the first major test publisher—founded byCattell, Thorndike, and Woodworth.First edition of the Strong Vocational Interest Blank published.Wechsler-Bellevue Intelligence Scale published. Revisions published in1955, 1981, and 1997.Minnesota Multiphasic Personality Inventory published.Wechsler Intelligence Scale for Children published. Revisions published in 1974, 1991.

CH01.QXD 6/12/2003 8:50 AM


1. In 1910, Henry Goddard translated the1908 Binet-Simon scale. In 1911, he tested morethan a thousand schoolchildren with the test, rely-ing upon the original French norms. He was dis-turbed to find that 3 percent of the sample was“feebleminded” and recommended segregationfrom society for these children.

2. Nonverbal intelligence tests were inventedin the early 1900s to facilitate testing of non-English-speaking immigrants. For example, Knoxpublished a wooden puzzle test in 1914 and alsoused the now familiar digit-symbol substitution test.

3. In 1916, Lewis Terman released the Stan-ford-Binet, a revision of the Binet scales. This well-designed and carefully normed test placedintelligence testing on a firm footing once and forall.

4. During WWI, Robert Yerkes headed a teamof psychologists who produced the Army Alpha, averbally loaded group test for average and superiorrecruits, and the Army Beta, a nonverbal group testfor illiterates and non-English-speaking recruits.

5. Early testing pioneers such as C. C.Brigham used results of individual and group in-telligence tests to substantiate ethnic differences inintelligence and thereby justify immigration re-strictions. Later, some of these testing pioneers dis-avowed their prior views.

6. Educational testing fell under the purviewof the College Entrance Examination Board(CEEB), founded at the turn of the twentieth cen-tury. In 1947, the CEEB was replaced by the Edu-

cational Testing Service (ETS), which supervisedthe release of such well-known tests as the Scholas-tic Aptitude Tests and the Graduate Record Exam.

7. The advent of multiple aptitude test batter-ies was made possible with the development of fac-tor analysis by L. L. Thurstone and others. Later,the improvement of these test batteries was spurredon by the practical need for selecting WWII recruitsfor highly specialized positions.

8. Personality testing began with Wood-worth’s Personal Data Sheet, a simple yes-nochecklist of symptoms used to screen WWI recruitsfor psychoneurosis. Many later inventories, includ-ing the popular Minnesota Multiphasic PersonalityInventory, borrowed content from the PersonalData Sheet.

9. Projective testing began with the word as-sociation technique pioneered by Francis Galtonand brought to fruition by C. G. Jung in 1910. Her-mann Rorschach published his famous inkblot testin 1921.

10. The Thematic Apperception Test (TAT), apicture storytelling test introduced in 1935 by Mor-gan and Murray, was based upon the projectivehypothesis: When responding to ambiguous or un-structured stimuli, examinees inadvertently dis-close their innermost needs, fantasies, and conflicts.

11. The assessment of vocational interestbegan with Yoakum’s Carnegie Interest Inventorydeveloped in 1919–1920. After several revisionsand extensions, this instrument emerged as E. K.Strong’s Vocational Interest Blank.

SUMMARY

CH01.QXD 6/12/2003 8:50 AM

Download - The History of Psychological Testing

Top Related