Leniency Error as a Function of the Rater\u27s Need for Social Approval

The effects of the rater\u27s need for social approval (N-SA) on leniency error in ratings was investigated. Each of 19 supervisors rated the performance of a common set of 12 workers. Using an ANOVA procedure (Guilford, 1954) leniency error was identified. The effects of N-SA as measured by the Personal Reactions Inventory (Crowne & Marlowe, 1960) and the Least Preferred Coworker (Fiedler, 1967) were partialed out of the ratings. When the ANOVA was repeated on the partialed ratings, it was hypothesized that reduced leniency error would result. Contrary to expectations, there was virtually no reduction in leniency error. The low reliability of the Personal Reactions Inventory, the questionable validity of the Least Preferred Coworker measure, and finally, the possibility that the underlying constructs of leniency error and N-SA are not related are offered as tentative explanations of this result. It was concluded, however, that this study may not have accurately measured the N-SA construct and directions for future research in this area are suggested