Looking Back and Looking Ahead in Psychometrics

Author(s):

Gulliksen, Harold O.

Publication Year:

1972

Report Number:

RM-72-08

Source:

ETS Research Memorandum

Document Type:

Report

Page Count:

30

Subject/Key Words:

History of Testing, Psychometrics, State of the Art Reviews

Abstract

A presentation of the 40-year history of psychometrics is given with comments about needed trends for the future. Computers have radically changed the time required for data processing. In testing, many promising developments, such as Kristof's reliability for vector variables, latent class and latent struction models, one-factor ration scale in testing and Bayesian procedures, are still largely in the theoretical field. Interest in scaling did not become important until Messick applied methods previously developed to attitude scales in 1956. Multidimensional scaling techniques have recently been utilized in a number of research areas and applied fields. Factor analysis theories are reasonably well developed. Applications to aptitude tests have been made, but have been only sketchily used in other fields in which they would be extremely valuable, such as economics, sociology, and physiology. In the field of mathematical learning theory, work needs to be done for individual learning curves and in comparing various stochastic and continuous models. Quantitative psychology has moved a long way in 40 years.

Read More

• Request Copy (specify title and report number, if any)