This paper examines how between-individual comparisons influence performance evaluations in rating tasks. We demonstrated a systematic change in the perceived difference across ratees as a result of changing the way performance information is expressed. Study 1 found that perceived performance difference between two individuals was greater when their objective performance levels were presented with small numbers (e.g., absence rates of 2% vs. 5%) than when they were presented with large numbers

more »

... (e.g., attendance rates of 98% vs. 95%). Extending this finding to situations involving trade-offs between multiple performance attributes across ratees, Study 2 showed that the relative preference for one ratee over another actually reversed when the presentation format of the performance information changed. We draw upon prospect theory (Kahneman & Tversky, 1979; Tversky & Kahneman, 1981) to offer a theoretical framework describing the between-individual comparison aspect of performance evaluation. APA IN PRESS ARTICLE: If you plan to quote this manuscript, be sure to check the actual wording of the final published article. 3 Between-individual comparisons in performance evaluation: A perspective from prospect theory One day, one of the authors overheard two students discussing the free throw performance of some NBA players. They tried to classify the players by giving them a grade from A+ to E. They mentioned Reggie Miller and agreed that he was an excellent free throw shooter and hence should be given an A grade. Then, they discussed Mike Bibby and had a disagreement. One said, "Mike's career free throw percentage is 80% and Reggie's percentage is 89%. Their performances do not differ much. I think Mike and Reggie should be classified with the same grade, though Reggie is better. So, I give Mike an A-." However, the other said, "No, Reggie missed only 11 free throws per 100. The miss percentage by Mike is 20%, which is almost double Reggie's. How can they be classified with the same grade? I think a B+ or even a B is more appropriate for Mike." The format of performance evaluation can be classified either as a relative rating system or an absolute rating system (see Cascio, 1998) . Relative rating systems ask raters to compare an employee's performance with the performance of other employees. Examples include simple ranking, paired comparisons, and forced distribution. Absolute rating systems ask raters to judge employees' performance on the basis of comparing their performance with performance standards, independent of between-individual comparisons. Examples include critical incidents, behavior checklists, and graphic rating scales. While making judgments about performance under absolute rating systems is supposed to be mainly determined by the discrepancy between the observed or recalled behavior and an absolute performance standard defined in subjective or objective terms (i.e., a behavior-standard