This study investigated the effects of accountability and purpose of appraisal on rating and behavioral accuracy. Subjects viewed a videotape of a lecture and were asked to rate the lecturer's performance. Accountability to the ratee (the GTA on the videotape) was held constant. Accountability to a supervisor (a faculty representative) was manipulated such that subjects in the no (supervisor) accountability condition anticipated a meeting with the GTA only; subjects in the weak (supervisor) accountability condition anticipated a meeting with the GTA as well as a supervisory review of the ratings; while subjects in the strong (supervisor) accountability condition were led to believe that they would have to meet with both the GTA and the faculty representative to explain their ratings. Additionally, participants were led to believe that the purpose of this appraisal was either to provide feedback for development or to make administrative decisions. Two-way ANOVAs were used to assess the effects of accountability and purpose of appraisal on rating accuracy (elevation accuracy, dimensional accuracy, leniency) and behavioral accuracy. Results indicated that (a) increased accountability to a supervisor led to greater elevation accuracy, (b) raters in the administrative purpose condition provided more lenient ratings than did raters in the developmental purpose condition , (c) behavioral accuracy increased with level of accountability (none, weak, strong) to a supervisor, (d) raters who believed that the purpose of appraisal was for development exhibited greater behavioral accuracy than did raters who believed that the purpose was to make administrative decisions. / Master of Science