The Effects of Imperfect Top-Down Selection on the Accuracy of the Direct Range Restriction Adjustment

Accurate assessment of the effectiveness of personnel psychology functions is vital to the field. Many personnel decisions are made based on correlations between predictor variables and measures of job performance; however, those correlations are often affected by range restriction. As encountered in applied practice, range restriction weakens the strength of the correlation. Equations exist to correct for the effects of range restriction on correlations; these equations are widely accepted and used by Industrial-Organizational psychologists today. This study expands on research by Hall (2016), which examined the accuracy of the direct range restriction correction equation provided by Thorndike (1949) under varying degrees of the violation of the assumption of perfect top-down selection. A Monte Carlo analysis was conducted so that the adjusted sample correlations could be compared to known, unrestricted population correlations. The results of the study indicate that Thorndike’s correction equation provides adjusted sample correlations that were closer to the true population correlation when perfect top-down selection does not occur. Implications and recommendations for future research are discussed