Abstract
The goal of the present study was to examine the effectiveness of a source-monitoring procedure in reducing halo error and negativity bias in a performance appraisal setting. Participants watched videotaped performance segments of an instructor presumably taking place over a school year (N = 126). Then, participants in the source-monitoring training condition received instructions teaching them how to differentiate memories that evoke detailed representations (remember judgments) from memories that evoke mere feelings of familiarity (know judgments). With this distinction in mind, participants in the training condition made significantly more accurate judgments than participants in the control condition. The training was successful in increasing the accuracy of performance ratings and reducing halo error and the negativity bias. Recognition memory measures showed that improvements were due to increased reliance on specific memory traces in the training condition. There was evidence of heuristic reliance on general impressions in the control condition.