Abstract The study is designed to test the impact of group composition variables on the supervisor's ratings of subordinates. In a laboratory study subjects play the role of supervisors evaluating the performance of their subordinates. The information about subordinate performance is constructed such that one subordinate performs poorly while two others perform well. The poor performer is represented as being high or low in group social skills and high or low in group leadership skills as rated by the other group members, creating a 2 × 2 design (the two good performers are portrayed as being average in social and leadership skills). The supervisor rates the poor performer more positively and the good performers more negatively when the poor performer is high in social and/or leadership skills than when he or she is low on these factors. A field study using critical incidents demonstrates that less severe responses are directed toward a popular group member. Implications of these findings for theory and practice are discussed.
[1]
Terence R. Mitchell,et al.
Effects of causal attributions on supervisor's evaluation of subordinate performance.
,
1980
.
[2]
Frank J. Landy,et al.
Statistical control of halo error in performance ratings.
,
1980
.
[3]
D. Kipnis,et al.
Untangling the performance appraisal dilemma: The influence of perceived organizational context on evaluative processes.
,
1976
.
[4]
T. Mitchell,et al.
Attributional processes of leaders in leader–member interactions.
,
1979
.
[5]
W. J. Bigoness,et al.
Effect of applicant's sex, race, and performance on employers' performance ratings: Some additional findings.
,
1976
.
[6]
Barton A. Weitz,et al.
Managing Marginal Employees: The Use of Warnings and Dismissals
,
1980
.