CONTEXT EFFECTS IN PERFORMANCE-APPRAISAL - A METHODOLOGICAL CRITIQUE AND EMPIRICAL-STUDY

The standard design used in research on assimilation and contrast effects in performance appraisal suffers from methodological flaws that preclude unambiguous interpretation of experimental results. This standard design is compared with two other designs that provide more appropriate tests of context effects (ef. Kravitz & Balzer, 1990). Undergraduates (N = 123) rated a videotaped lecture of average quality after rating (a) two videotapes depicting good lectures, (b) two videotapes depicting poor lectures, or (c) no other videotapes. Half the subjects had rated the target videotape 1 week earlier. Analysis of the standard design implied contrast effects. Analyses of the alternative designs revealed problems with the positive context manipulation, assimilation effects in the positive context, contrast effects in the negative context, and pretest effects. It is recommended that the standard design not be used.