F. L. Schmidt et al argued that the average (median sample size 68) validity study has a 50% chance of detecting validity if validity exists. They assumed direct range restriction (RR) in determining the sample size needed for adequate power. Because indirect RR is more typical, 4 tables indicating required sample sizes under indirect RR are presented. Contrary to Schmidt et al, sample sizes needed under indirect RR are substantially smaller than those needed under direct RR. Under indirect RR, the average validity study has a 75% chance of detecting validity if validity exists when a 1-tailed test can be used. (13 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
- computer program used to develop tables of required sample size under indirect range restriction, criterion-related validity, extension of work of F. L. Schmidt et al