The Psychometric Society

The Psychometric Society is an international non-profit professional organisation devoted to the advancement of quantitative measurement practices in psychology, education and the social sciences.

The Society publishes the journal Psychometrika which contains articles on the development of quantitative models of psychological phenomena, as well as statistical methods and mathematical techniques for evaluating psychological and educational data.

History of psychometrics

Psychometrics began in the late 19th century in Europe through the joint  influences of the anthropologist Francis Galton in London and the psychophysicist Wilhelm Wundt in Leipzig, Germany. The world's first psychometric laboratory was established in Cambridge in 1886 by James McKeen Cattell, an American collaborator of Galton who had completed his PhD with Wundt. After three years in Cambridge, Cattell took up a Chair in Psychology at the University of Pennsylvania in 1889, and moved to Columbia University in New York a few years later where he became a pivotol figure in the founding of American psychology. Galton's legacy was the establishment of a department at University College London that saw the development of many of the pioneering ideas in statistics by Karl Pearson, Charles Spearmen, R A Fisher and others.

Psychometrics in the USA

The London group's primary interest was genetics and anthropometrics, with an emphisis on a single trait of intelligence now generally referred to simply as Spearman's "g", and their thinking influenced some of the early American psychometricians such as Terman and Yerkes.  However, many felt this to be an oversimplification and alternative views of mental measurement soon begun to have an impact. In 1935 some of the most important psychometricians of the time (Thurstone, Thorndike, Gulliksen, Guilford, and McNemar) banded together to form the Psychometric Society and establish the journal Psychometrika.

The founding of the Psychometric Society

In 1936 L.L. Thurstone, Professor of Psychology at the University of Chicago, became the first President of the newly formed Psychometric Society. Today he is remembered for his contribution to scaling techniques in attitude measurement, and for his theory of primary mental abilities. Thurstone argued, contra Spearman, that there was more than one form of intelligence (Psychological Review, 1934, 41, 1-32). His primary abilities were: word fluency, verbal comprehension, spatial visualisation, number facility, associative memory, reasoning and perceptual speed. He was able to show that different forms of factor analysis could produce different solutions, and believed that Spearman’s single factor, or “g”, was an artefact of the method used. Thurstone laid the foundation for the differential aptitude tests still in use today.

Edward L. Thorndike, the second president of the Psychometric Society, obtained his PhD at Columbia in 1898 under the supervision of James McKeen Cattell.Thorndike was primarily an applied psychometrician. His CAVD (completion, arithmetic, vocabulary, and directions test) developed in the 1920s contained a fundamental design that became the foundation of modern intelligence tests. He distinguished three forms of intelligence; abstract intelligence, mechanical intelligence, and social intelligence, and hence, like Thurstone, was also at odds with Spearman, as well as anticipating many of the forms of multiple intelligence that became popular 50 years later.

J. P. Guilford followed as President of the Society in 1938. His Structure of Intellect Model saw intelligence as a number of abilities for processing different kinds of information. There were six kinds of operations (cognition, memory recording, memory retention, convergent production, divergent production, and evaluation); five kinds of contents (visual, auditory,symbolic, semantic, and behavioral); and six kinds of products (units, classes, relations, systems, transformations, and implications). Altogether these yielded 180 possible abilities. His extension of Thurstone's and Thorndike's several intelligences to more than 100 had a profound impact on the way in which intelligence was seen by psychologists and lead to an explosion in new and varied approaches to its measurement.

Harold Gulliksen, the 9th President, was Professor of Psychology at Princeton University and a research adviser on testing to many government organisations.  His book ''The Theory of Mental Tests,'' had a major impact on students of assessment and remained in print for many decades. When the Educational Testing Service was founded in 1948, he directed its psychometric fellowship program. From 1942 to 1949 he was managing editor of Psychometrika.

Quinn McNemar was President of the Psychometric Society in 1951. For many years McNamar was Professor of Psychology, Statistics, and Education, at Stanford University. He had more training in mathematics than in psychology, so that statistics became a natural emphasis, although it was essentially a new field for him.

In 1929, after hearing Charles Spearman talk at the International Congress of Psychology at Yale University, he visited Thurstone at the University of Chicago where the alternative positions of factor analysis were presented in person by Spearman, Holzinger, Thurstone, and Hotelling. Then at Columbia University during the remainder of the year he learned of a new technique called analysis of variance, recently introduced by R. A. Fisher from England. He was aided in this by studying with Harold Hotelling, who had traveled to Oxford to get the story directly from Fisher himself. The first edition of McNemar's influential book ‘Psychological Statistics’ appeared in 1949, to be followed by revisions in 1955, 1962, and 1969.

Lee J Cronbach, President of the Society in 1953 originated many classical psychometric formula, including a measure of the reliability of a psychological or educational test known as "Cronbach's alpha". He also developed a theory of test reliability and validity known as "Generalizability Theory", a comprehensive statistical model for the identification of sources of measurement error. When he served as President of the American Psychological Association in 1956­/57 he sought to bridge the gap between different theories of psychology by showing the importance of both the environment and individual behavior.

Many other distinguished contributions to psychometrics were made during these early years by other Presidents: Truman Kelly, Karl Holzinger, Jack Dunlap, Paul Horst, Marion Richardson, Henry Garrett, Edward Cureton, Harold Edgerton, Irvine Lorge, Phillip Rulon, Dorathy Adkins, John Flanagan, Robert L. Thorndike, Ledyard Tucker, Clyde Coombs, Hubert Brogden and Frederick Mosteller.

The modern era in psychometrics

The modern era in Psychometrics began with the publication in 1969 of the seminal work "The Theory of Mental Test Scores" by Frederic Lord and Melvin Novick. Frederic Lord had been President of the Society in 1958. and the listing of Past Presidents since then is a roll call of fame, not limited to psychometrics alone. Many, indeed most, have had an influence far beyond; on developments in statistics, computing, classification, psychology, social science, education and medicine. They are Lloyd Humphreys, John Carroll, Philip Dubois, Lyle Jones, Allen Edwards, Warren Torgerson, Bert Green, Chester Harris, B J Winer, Harry Harman, Henry Kaiser, Louis Guttman, Samual Messick, Darrell Bock, Roger Shepard, Joseph Kruskal, Douglas Carroll, Duncan Luce, Karl Joreskog, Norman Cliff, Melvin Novick, Forrest Young, James Ramsey, Peter Bentler, Larry Hubert, Bruce Bloxom, Roderick McDonald, Yoshio Takane, Jan de Leeuw, Bengt Muthen, Paul Holland, Phipps Arabie, Michael Browne, William Meredith, Robery Mislevy, Gerhard Fisher, Shizuhiko Nishisato, Fumiko Samejima, Ivo Molenaar, Susan Embretson, Wim van der Linden, David Thissen, William Stout, Jacqueline Meulman, William Heiser, Robert Cudeck, Ulf Bokenholt, Roger Millsap and Paul de Boeck. The current president of the Psychometric Society is Brian Junker.