EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT THERON THE EFFECT OF A JOINT CORRECTION FOR THE ATTENUATING EFFECT OF CRITERION UNRELIABILITY AND CASE 1 RESTRICTION OF RANGE ON EXPECTED CRITERION PERFORMANCE AND THE CONDITIONAL PROBABILITY OF SUCCESS

This article reports the results of a portion of a more comprehensive study on the effect of corrections for random error of measurement and/or various forms of restriction of range on those estimates required to specify and justify a selection procedure. The objective of this article is to determine the effect of a joint correction for criterion unreliability and Case 1 selection on the predictor on the expected criterion performance, conditional cri-terion variance, and selection decision making. Personnel selection is necessitated by the combined effect of interindivid-ual differences among applicants on those attributes that would determine their eventual job performance and the selecting organization’s desire to maximize performance. The desire to maximize performance implies work success as the ultimate/final institutional criterion in terms of which appli-cants for employment should ideally be evaluated and on which they should ideally be compared so as to arrive at an institutionally rational selection de-cision. Personnel selection is, however, complicated by the obvious fact that information on the ultimate institutional criterion can never be available at the time of the selection decision. The only solution to this dilemma, apart from reducing selection to random assignment, is to base the decision on rele-vant substitute information that is assessable prior to the selection decision. Even if no direct information on the criterion ever enters the selection decision-making process, the criterion nonetheless always remains the focus of interest in selection assessment. This seemingly innocent and too often forgotten fact, moreover, has important implications for the interpretatio