Doctoral Dissertations

Date of Award

5-1992

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Industrial and Organizational Psychology

Major Professor

Gregory H. Hobbins

Committee Members

Tom Ladd, Joyce Ruseell, John Philpot

Abstract

Although ratings of previous performance have been shown to bias ratings of subsequent performance in the performance appraisal literature, the effect of previous performance information has not been examined in assessment center research. The primary purpose of this project was to determine if the effects of previous performance described in the performance appraisal literature would generalize to assessment center ratings. This research also attempted to show that decreasing the cognitive demands on assessors would result in ratings which were less subject to previous performance effects. Since assessors often rate multiple applicants on a number of different performance dimensions, cognitive demand was manipulated by varying both the number of dimensions and the number of applicants rated. The research consisted of a laboratory experiment and a field study. The participants for study 1, the laboratory experiment, consisted of 100 management and psychology students at a large southeastern university. The participants received 1.5 hours of assessor training and were randomly assigned to one of eight experimental conditions. Assessors in the high performance condition rated an applicant exhibiting good initial performance, while assessors in the low performance condition rated an applicant exhibiting poor initial performance. Assessors in the high cognitive demand condition rated eight performance dimensions, and assessors in the low demand condition rated only four dimensions. Cognitive demand was also manipulated by having assessors rate two applicants in the leaderless group discussion and one applicant in the interview exercise. For the subsequent exercises, participants in all eight groups observed and rated identical videotapes of the same applicant portraying average performance. The results of study 1 confirmed the hypothesis that assessors who observed high performance on the initial exercise rated the applicant higher on subsequent exercises than assessors who observed low performance on the initial exercise. Since the applicant's true performance was identical for all assessors, these results suggest that assessors' knowledge of previous performance biased their ratings of subsequent exercises. The results of study 1 provided some support for the hypothesis that previous performance bias is more likely to occur when the information processing demand on assessors is high. In fact, the ratings made by assessors in the low cognitive demand condition were not significantly affected by ratings of previous performance. However, study 1 provided little support for the hypothesis that having to rate an additional applicant would result in more biased ratings. The data for study 2 were collected as part of a concurrent validation of a customer contact assessment center at a major corporation. Eleven customer service managers received three full days of assessment center training and were assigned to observe and rate the performance of 84 randomly selected customer service employees. Study 2 provided a limited test of the generalizability of previous performance effects to a field setting and also examined the effects of previous performance bias on the validity of assessment center ratings. The results of study 2 provided some support for the hypothesis that previous performance bias may have influenced assessors' ratings of the customer service assessment center. However, the hypothesis that ratings of subsequent exercises would not add incremental variance to the prediction of job performance was rejected. In summary, this research marks one of the first attempts to examine the assessment center from an information processing perspective. The research demonstrated that previous performance bias can occur in an assessment center context, even when the exercises are quite different in format. There were also some indications that previous performance effects are more likely to occur when increased cognitive demands are placed on the assessors and that previous performance effects may occur in field settings using managerial assessors, implications of these results and directions for future research are also discussed.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS