Effects of processing method, performance pattern, and time pressure on performance ratings
Spychalski, Annette C.
Doctor of Philosophy
One hundred seventy-eight subjects observed videotaped incidents of a secretary performing on the job and rated her performance across four performance dimensions (and overall performance). Rating accuracy, behavior recognition accuracy, and rating level were measured under conditions of viewing improving or deteriorating performance, with behavior-based or impression-based processing methods, and with or without time pressure to complete the rating task. Results display robust main effects for both processing method and performance pattern on rating level. Raters viewing performance that improved over time gave the secretary lower ratings than raters viewing performance that deteriorated over time. Raters using behavior-based processing methods rated performance higher than raters using impression-based processing methods during the rating task. Processing method influenced rating accuracy for two of the performance dimensions; effects were mixed. Small relationships among measures of memory discrimination and response bias and rating accuracy and performance pattern were also discovered. Time pressure had no stable effect on any of the dependent variables. The results illustrate the pervasiveness of primacy effects in performance rating tasks. They also underscore the importance of standardizing rating procedures that are used to compare performance of different individuals. Additional research in applied settings is needed to capture the organizational influences on the performance rating process. Because of the complexities involved in performance rating systems in organizations, it may be prudent to change the typical way in which supervisor input about employee performance is used in human resources decisions.