The “coefficient of forecasting efficiency,” known as E, has been rejected as a measure of test value by decision theorists. A common type of test-use situation is described and a function is derived measuring test value in this type of situation. It is shown that in certain instances this function reduces to E, so that E is sometimes an appropriate measure of test value. The article describes this special instance first, and then the more general formula is derived.
Get full access to this article
View all access options for this article.
References
1.
BrogdenH. E.A new coefficient: Application to biserial correlation and to estimation of selective efficiency. Psychometrika, 1949, 14, 169–182.
2.
CronbachL.Essentials of psychological testing.New York: Harper, 1960.
3.
CronbachL.GleserG. C.Psychological tests and personnel decisions.Urbana: Univer. of Illinois, 1957.
4.
HullC. L.Aptitude testing.Yonkers-on-Hudson: World Book, 1928.
TaylorH.RussellJ.The relationship of validity coefficients to the practical effectiveness of tests in selection: Discussion and tables. J. appl. Psychol., 1939, 23, 565–578.
