* Academy of Management Journal 1986, Vol. 29, No. 1, 115-129. EFFECTS OF GENDER ON SELF- AND SUPERVISORY RATINGS LYNN McFARLANE SHORE GEORGE C. THORNTON III Colorado State University This research investigated the effects of supervisors' and suhordinates' genders on self- and supervisory ratings in an organizational setting. Participants were assemblers, 35 men and 35 women, and their super- visors, 16 men and 19 women. Results showed that subordinates' self- ratings were higher than their supervisors' ratings of them and that gender did not affect the relationship between self- and supervisory ratings. In prior research involving novel situations in laboratory settings, women provided lower self-ratings and higher ratings of others than men. The present study showed no gender differences in ratings on familiar tasks in a real work setting in which performance feedback was available. Concern for minimizing error in ratings has produced many and varied approaches to investigating accuracy in performance appraisals (Landy & Farr, 1980). The present study synthesized two distinct research approaches to accuracy and fairness in performance evaluations. First, much research has compared self- and supervisory ratings in work settings in an effort to evaluate the accuracy of self-assessments for organizational purposes (Thornton, 1980). Second, concern about fair evaluations of men and women has led to extensive laboratory research on gender effects in evaluation situa- tions (Nieva & Gutek, 1980). The present study integrated these two areas of research to address issues of accuracy raised by both literatures. Theory and selected research results suggest that previous laboratory findings about gender differences in ratings may have limited generalizahility to the workplace. Therefore, the present study investigated gender differ- ences in a realistic setting. Men and women employees and supervisors in an actual work organization participated in the study. REVIEW OF LITERATURE AND HYPOTHESES Congruence between Self- and Supervisory Ratings Many researchers have suggested that self-ratings serve a variety of organi- zational purposes (Meyer, 1977; Primoff, 1980). The use of self-ratings has We gratefully acknowledge the assistance of James McCambridge for his contributions to the development of this research project; Alan Inada for his help in data collection; and Ted H. Shore for his editorial assistance. 115