Corrections for restriction of range and attenuation in criterion-related validation studies.

The results of 1,253 Navy enlisted personnel on the Navy Basic Test Battery and the Navy Enlisted Advancement Exam were used as predictor and criterion, respectively. Groups were selected according to 5 selection ratios, .10 to .50. Results were then compared to the true validity. Except for the selection ratio of .10, all of the corrected validity coefficients slightly over-estimated the true validity. Triple and 2-step correction procedures yielded equivalent results. In all cases, the corrected coefficients better estimated the unrestricted true validity. (4 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)