Open AccessTHE USEFULNESS OF SELECTION TESTS IN COLLEGE ADMISSIONS
Author(s) -
Beaton Albert E.,
Barone John L.
Publication year - 1981
Publication title -
ets research report series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.235
H-Index - 5
ISSN - 2330-8516
DOI - 10.1002/j.2333-8504.1981.tb01247.x
Subject(s) - selection (genetic algorithm) , statistics , test (biology) , mathematics , interpretation (philosophy) , variable (mathematics) , econometrics , psychology , computer science , artificial intelligence , paleontology , mathematical analysis , biology , programming language
In many selection situations, some predictor of success, such as a test, is used to discriminate between the applicants to be accepted and those to be rejected. Using an interpretation of Brogden [1946], this paper shows the effect of using a correlated variable on the mean criterion score of the selected group. The distribution of the means of all possible selections is discussed and a simple formula is provided for calculating the probability of achieving as high or higher mean using random selection as opposed to using a correlated variable. It is shown that even fairly small correlations result in selection decisions that are not often exceeded by chance. The Brogden interpretation and significance test is shown to extend simply to the case of several predictors.