Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
A personal construct model of performance appraisal
Abstract
A model of performance appraisal based on George Kelly's (1955) personal construct theory is proposed. Much of the current literature on cognitive models of performance appraisal is reviewed. Research in the past has dealt primarily with changes in rating scale format and rater training. Limited success has been achieved in reducing psychometric errors and increasing the accuracy of performance ratings. The present model is proposed to account for much of the idiosyncratic variance in ratings which is often considered "error" variance. It suggests that a portion of observed errors and inaccuracies in performance ratings is due to the failure of the performance appraisal process to tap into the personal constructs of the rater. Three studies tested the applicability of the personal construct model. The first study examined the differential reliability of ratings made on personal constructs and provided scales. The data suggests personal construct ratings to be measurable consistent and more reliable than provided construct ratings. The second study examined the recognition of performance related behaviors after having the encoding of these behaviors primed with either personal or provided construct dimensions. The findings for this study indicate personal construct encoding increases positive and negative behaviour recognition while decreasing false positives in comparison to provided construct encoding. The third study examined the proportion of overall effectiveness variance which can be predicted by ratings on personal and provided construct scales. It was found that personal construct ratings account for a significantly greater amount of variance in overall ratings. In addition, the psychometric rating errors of halo and leniency were examined for both types of scales. The results indicated no difference in the occurrence of psychometric errors between ratings on personal and provided constructs. Potential uses for the model such as rater training and appraiser/appraisee feedback conferences are discussed.
Subject Area
Occupational psychology|Psychological tests
Recommended Citation
Downs, Alan Wilson, "A personal construct model of performance appraisal" (1989). ETD collection for University of Nebraska-Lincoln. AAI8925233.
https://digitalcommons.unl.edu/dissertations/AAI8925233