An examination of the effects of rater training and individual prototypes on performance ratings

Abstract

The performance appraisal process is often inaccurate due to various biases and psychometric errors. Rater training was developed in an attempt to familiarize raters with the dimensions used to judge performance and to standardize the appraisal process. It was expected that as a result of training, raters would develop a common frame-of-reference on which to rely during ratings. The goals of the present study were to validate the effectiveness of a common rater-training program, Performance Dimension Training, and to examine the effects of Performance Dimension Training on individual prototypes. Participants in the training condition were trained on the multidimensionality of professor performance and on the specific dimensions and behaviors that constitute effective performance. It was hypothesized that subjects in the training condition would produce more accurate performance ratings. It was also believed that individual's prototypes of professor performance would change to conform toward the"'ideal""prototype presented as the standard during training. As a result individuals would rely on this prototype standard when completing performance ratings. This study was also concerned with under what conditions Performance Dimension Training is effective. Results showed a main effect for training on accuracy as training conditions were more accurate on performance ratings. A main effect was also found for the multi-tasking manipulation with subjects in the multi-tasking condition making less overall accurate performance ratings. Subjects in the training conditions did change their prototypes more so then did those in the control conditions, as hypothesized. Discussion and implications concerning these findings are presented.