ChapterPDF Available

Cronbach, Lee J. (1916–2001)



Lee J. Cronbach (1916–2001) had a long and distinguished career as an educational psychologist. He made major contributions to psychological testing, including coefficient alpha, generalizability theory, and construct validity theory; to evaluation, including a focus on program improvement, an emphasis on the role played by context and politics, and the strengths and limitations of randomized trials; and to instruction, most notably in the search for aptitude-treatment interactions and designing particular learning environments to match learners' characteristics.
360 May 2002 • American Psychologist
Copyright 2002 by the American Psychological Association, Inc. 0003-066X/02/$5.00
Vol. 57, No. 5, 360–361 DOI: 10.1037//0003-066X.57.5.360
Lee Joseph Cronbach made major contributions in the fields of
educational psychology, psychological testing, and program
evaluation throughout a career that spanned more than five
decades. Harbingers of his career in testing were evident at an
early age in Fresno, California, where he was born on April 22,
1916, to a homemaker and salesman. According to his sister,
Lee was overheard at age 4 years in a grocery market calculat-
ing the unit price of potatoes, drawing the conclusion that the
market his mother shopped at charged far more than did the
market of his babysitter. The feat was reported to Blanche
Cummings, a school psychologist and disciple of Lewis Terman,
who gave Lee an IQ test in 1921 and enrolled him in Terman’s
longitudinal study of gifted children. Pushed through school
by his mother, Lee graduated from Fresno High School at 14
years of age and Fresno State Teachers College (majoring in
chemistry and mathematics) at 18.
Lee’s awareness of Terman and IQ testing attracted him
to psychology, and, as a college junior, he came across a
monograph by L. L. Thurstone and E. J. Chave, The Measure-
ment of Social Attitudes (1931). The connection between psy-
chology and education came in a course at Fresno State on the
history of American education. Asked to review and present a
curriculum-improvement study to his fellow students, he dis-
sected the study with the scorn of a chemistry student. After the
presentation, the professor asked if he had thought of a career
in education research; Lee asked if there were such a thing. The
suggestion came at a time when he had decided against
pursuing chemistry, and the path was set toward educa-
tional psychology.
Lee taught high school from 1936 to 1938 while earning
a master’s degree from the University of California, Berkeley
(1938). He then completed a doctorate in education at the
University of Chicago (1940). During this period, Lee married,
and he and his wife, Helen, had five children between 1941
and 1956.
Beginning in 1940, Lee held a number of academic posi-
tions before finally settling at Stanford University in 1964,
where he became the Vida Jacks Professor of Education. From
1940 to 1946, he was an assistant professor at Washington
State University. Toward the end of World War II, he served as
a military psychologist at the U.S. Navy’s sonar school in San
Diego, California. Lee was an assistant professor at the Univer-
sity of Chicago from 1946 to 1948, then professor at the
University of Illinois from 1948 to 1964. During this period, he
also served as the Office of Naval Research’s science liaison in
London (1955–1956), as a member of Princeton’s Institute for
Advanced Study (1960–1961), and as a fellow at the Center for
Advanced Study in the Behavioral Sciences in Palo Alto, Cali-
fornia (1963–1964).
Lee’s research can be clustered into three areas: mea-
surement theory, program evaluation, and instruction. His most
widely cited measurement article was “Coefficient Alpha and
the Internal Structure of Tests” (1951). The coefficient, known
as Cronbach’s alpha, proved useful for (at least) three reasons.
First, it provided a measure of reliability from a single test
administration so that repeated occasions or parallel forms of a
test were not needed to estimate a test’s consistency. Second,
the formula was general; it could be applied to dichotomously
scored multiple-choice items or polytomous attitude scales.
And third, it was very easily calculated from statistics well-
known by students with only a first course in statistics.
A book Lee wrote with Goldine Gleser, The Dependabil-
ity of Behavioral Measurements (1972), addressed some of the
issues he believed were crucial in measurement theory. He and
Gleser revealed the underlying assumptions of the various
approaches to reliability and how they related to each other.
They showed how one could design an experimental proce-
dure and analyze data obtained thereby using a random model
analysis of variance to estimate the contributions to the error
variance from various sources for a specified true score. This
information could then be used to develop an improved testing
procedure for decision making. What started out to be a hand-
book on measurement with Gleser became a major
reconceptualization of reliability theory in the form of
generalizability theory, or G theory. G theory provided a meld-
ing of the psychological with the mathematical and produced a
comprehensive framework and statistical model for identify-
ing sources of measurement error.
Lee’s work on validity theory—the extent to which an
interpretation of a test is conceptually and empirically war-
ranted—was no less significant. With Paul Meehl, he placed
this idea at the center of psychological, educational, and social
testing. Validation, a never-ending process, examined a pro-
posed test interpretation—a construct—by testing it logically
and empirically against counterinterpretations. Moreover, what
was validated, according to Lee, was not the test itself, for a test
could be used for many purposes (e.g., prediction, diagnosis,
or placement). Rather, what was validated was a proposed
interpretation. Moreover, although reliability was an impor-
tant characteristic of a test, Lee believed that ultimately reli-
ability served its master, validity, where sometimes trade-offs
were necessary between broadly gauging a construct and nar-
rowly constructing a homogeneous set of items to improve
reliability. The Cronbach and Meehl (1955) article “Construct
Validity in Psychological Tests” laid the groundwork for fur-
ther work on validity.
Lee was skeptical of the sterile view of evaluation as
detached, objective, scientific activities with test content
matched to curricula, appropriate experimental designs, and
proper statistical tests. In the 1970s, he directed the Stanford
Evaluation Consortium, a research, service, and training orga-
nization sponsored by the Stanford University School of Edu-
cation that included faculty from the education, communica-
tions, and psychology departments. The consortium worked
Lee J. Cronbach (1916–2001)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
May 2002 • American Psychologist 361
with the state of California to examine its relationships with
local school districts, among other projects. The evaluation
research influenced program evaluations across many fields,
from health programs to juvenile delinquency programs; his
work recognized the merits and limitations of randomized
field trials, the importance of local contexts to performance,
and the social and political aspects of program evaluation.
Two influential books resulted from this work: Towards Re-
form of Program Evaluation (1980), which was written by a
team of consortium faculty led by Lee, and a parallel volume,
Designing Educational Evaluations (1982), in which Lee ex-
tended his own ideas.
Lee’s instructional research with Dick Snow focused on
matching learning environments with students’ aptitudes
(Cronbach & Snow, Aptitudes and Instructional Methods,
1977). This research can be traced to early work with Goldine
Gleser, first published as Psychological Tests and Personnel
Decisions (1957), and his presidential address to the American
Psychological Association (APA; 1957). In the work on per-
sonnel placement, he and Goldine concluded that optimal
decisions about person–job matches must acknowledge the
interaction of individual differences with job demands. Indi-
viduals with one profile of characteristics would be expected
to perform well in one type of job, whereas individuals with a
second profile would be expected to perform well in another
job with different task demands.
The work on personnel theory provided a fresh look at
the schism in scientific psychology, one that formed the basis
of Lee’s presidential address to the APA, “The Two Disci-
plines of Scientific Psychology” (Cronbach, 1957). He called
for a rapprochement between the two major investigative ap-
proaches of psychologists, the correlational work that was
most characteristic of individual-differences research and the
experimental work that focused on differences between situa-
tions or treatments. Lee pointed out that these individual dif-
ferences might be highly predictive of performance in one
type of instructional condition (situation) and much less so in
another. If individual-difference and experimental psychol-
ogy came together, he reasoned, it just might be possible to
find a link between these individual differences and perfor-
mance in different learning environments.
Lee’s work on personnel decisions and his presidential
address sent him and Dick Snow on a 10-year trek in search of
aptitude–treatment interactions (ATIs) or statistical interac-
tions (different regressions of learning outcome on aptitude
under different instructional treatments). The strongest ATIs
involved general ability, where students with above-average
intellectual development profited from instruction that pro-
vided them with considerable responsibility for organizing
and interpreting, whereas those below average profited from a
highly structured learning environment. In 1975, Lee revisited
his presidential address in the article “Beyond the Two Disci-
plines of Scientific Psychology,” where he revealed that he
had come to recognize that ATIs were highly complex—as if a
simple ATI were reflected in a hall of mirrors—sometimes
rapidly changing, and context bound, far more than he had
imagined earlier on. He concluded that “our troubles do not
arise because human events are in principle unlawful; man and
his creations are part of the natural world. The trouble, as I see
it, is that we cannot store up generalizations and constructs for
ultimate assembly into a network” (Cronbach, 1975, p. 123,
italics in original).
In the end, Lee found himself caught between science
and practice, whether in the classroom or in policy. Science
took him just so far, and he demanded science as far as it would
take him. But he also recognized the contributions that other
ways of knowing had to make in understanding teaching and
learning, as well as human action more generally:
The special task of the social scientist in each generation is to pin
down the contemporary facts. Beyond that, he shares with the
humanistic scholar and the artist in the effort to gain insight into
contemporary relationships, and to align the culture’s view of man
with present realities. To know man as he is is no mean aspiration.
(Cronbach, 1975, p. 126)
It was not for acclaim that Lee took on most offices but
rather out of a deep sense of responsibility, as well as the belief
that he had something to add. His contributions to APA were
many. He chaired the first committee on test standards from
1950 to 1953. These standards were endorsed by the American
Educational Research Association and the National Council
on Measurement in Education and resulted in a collaboration
that sponsored all subsequent revisions of test standards. From
1951 to 1953, Lee served as chair of the Publications Board.
He was on the Board of Directors from 1952 to 1958 and was
president of the association in 1956.
Lee’s professional honors were numerous. He was presi-
dent of the American Educational Research Association and
the Psychometric Society. Lee was elected to membership in
the National Academy of Sciences, the National Academy of
Education, the American Philosophical Society, and the
American Academy of Arts and Sciences. He received many
honorary degrees, including ones from Yeshiva University;
the University of Gothenburg, Sweden; and the University of
Chicago. And he was honored by, for example, the Educa-
tional Testing Service for contributions to educational mea-
surement, the APA for distinguished scientific contributions,
the American Psychological Society as a William James Fel-
low, the American Educational Research Association for con-
tributions to research in education, and the Evaluation Re-
search Society for contributions to evaluation methodology.
Lee retired in 1980 but remained intellectually active
right up until the time of his death. For example, he was writing
an article on the 50th anniversary of the coefficient alpha
paper, in which he reflected on the uses, misuses, and misun-
derstandings of the reliability coefficient—an article that, un-
fortunately, will never be finished.
Lee J. Cronbach died of congestive heart failure in his
Palo Alto home on October 1, 2001. Lee’s wife, Helen, and
three of his five children—Janet, Bob, and Joyce—survive
him, as does his sister, also named Helen.
Richard J. Shavelson
Stanford University
Goldine Gleser
University of Cincinnati
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
In 1997, noting that the 50th anniversary of the publication of “Coefficient Alpha and the Internal Structure of Tests” was fast approaching, Lee Cronbach planned what have become the notes published here. His aimwas to point out theways in which his views on coefficient alpha had evolved, doubting nowthat the coefficientwas the bestway of judging the reliability of an instrument to which it was applied. Tracing in these notes, in vintage Cronbach style, his thinking before, during, and after the publication of the alpha paper, his “current thoughts” on coefficient alpha are that alpha covers only a small perspective of the range of measurement uses for which reliability information is needed and that it should be viewed within a much larger system of reliability analysis, generalizability theory.
General methodological difficulties are discussed, particularly; the need to discuss similarity only with respect to specified dimensions, loss of information involved when configurations are reduced to indices, the need to interpret a similarity index as a relative rather than as an absolute measure, and the general non-comparability of scale units involved in profiles. The measure D is presented. This is, for two profiles, the sum of the squared deviations of corresponding scores, and is a general expression for dissimilarity (distance in the hyperspace of k variates). 27 references.
Examines procedures previously recommended by various authors for the estimation of "change" scores, "residual," or "basefree" measures of change, and other kinds of difference scores. A procedure proposed by F. M. Lord is extended to obtain more precise estimates, and an alternative to the L. R. Tucker, F. Damarin, and S. A. Messick (see 41:3) procedure is offered. A consideration of the purposes for which change measures have been sought in the past leads to a series of recommended procedures which solve research and personnel-decision problems without estimation of change scores for individuals. (22 ref.) (PsycINFO Database Record (c) 2012 APA, all rights reserved)