Effects of Subsequent Performance on Evaluations of Previous Performance

It was hypothesized that subsequent performance levels would bias the recall and evaluations of a ratee's previous level of performance. Subjects (N = 183) rated three videotaped lectures in either immediate or delayed rating conditions. The first videotape depicted an average level of performance and was followed by either two good lectures or two poor lectures. A significant Performance Level × Time of Rating interaction was found, in which memory-based ratings were biased in the direction of subsequent performance. That is, when there was a delay between observation and rating, subjects who had seen an average lecture followed by good lectures rated that average lecture more favorably than did subjects who had seen that same lecture, followed by poor lectures. We suggest that raters are biased in favor of recalling behaviors that are consistent with their general impression of a ratee and that subsequent performance may systematically alter the rater's recall of the ratee's previous behavior.