Validity of selection tests: a study

Abstract

Several reasons have been offered for the depressed values of coefficients of correlation between performance evaluation scores and test scores for tests that otherwise seem to have high validity. Most of these studies have concerned themselves with only the first year performance measure. This study was undertaken to broadbase the validity design by including performance measures of three subsequent years. Data on the test and performance scores of a sample of executives were analysed. The results indicate that though test scores may not show significant relationship with the first year performance appraisal score, they show positive and significant relationship with subsequent performance appraisal scores. The results are explained in terms of changed performance evaluation.