Advisor(s) - Committee Chair
John O'Connor, Daniel Roenker, Carl Martray
Department of Psychology
Master of Arts
The major purpose of the study was to describe, analyze, and compare individual performance rating behavior in terms of the criteria used and the relative importance of the criteria. In order to objectively describe individual rating behavior, a policy capturing procedure was used. Nineteen faculty and the chairperson in the psychology department of a regional state university participated in the study. Using a behavioral expectation scaling methodology, 14 performance dimensions were identified and anchored. One hundred hypothetical performance profiles were generated, with scores from one to nine on each of the 14 dimensions. Subjects reviewed each profile and assigned an overall rating. Additionally, they directly estimated the relative importance of each of the 14 dimensions in a) their own rating policy and b) the chairperson’s rating policy. For each subject, overall ratings were regressed on the 14 dimensions. Also, an index of the relative importance of each of the individual dimensions was computed. A hierarchical clustering procedure grouped subjects according to similarities in their regression equations. Discrepancy indices were computed between a) each subject’s estimated and statistical dimension weights and b) each subject’s estimation of the chairperson’s dimension weights and his statistical weights. Results showed that subjects were very consistent in making overall ratings (median R2 = .77), yet evidenced substantial variation in the use of different performance dimensions. The clustering procedure yielded a four group solution. Composite cluster rating policies revealed a number of important and interpretable differences in terms of dimensional weighting. On average, subjects reported that 13 of the 14 dimensions were important in their own policies, while only three dimensions were needed to account for over 71% of the predictable variance. Faculty members also overestimated the number of dimensions used by the chairperson (on average, 13.3), while four accounted for 79% of his predictable variance. In summary, the policy capturing methodology provides a means of accurately and objectively describing the criteria used in rating behavior. Implications suggest that by providing a supervisor’s rating policy to subordinates, one could clarify what aspects of performance are considered to be most important. Additionally, capturing the rating policies of both subordinates and supervisor can provide a framework within which a rational discussion of policy differences can proceed, with the distinct opportunity for the development of individualized rating contracts.
Business | Human Resources Management | Industrial and Organizational Psychology | Performance Management | Psychology | Social and Behavioral Sciences
Hobson, Charles, "Clarifying Performance Appraisal Criteria" (1980). Masters Theses & Specialist Projects. Paper 3293.