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.
Influence of assessor individual differences on rating errors and rating accuracy in assessment centers
The predictive influence of assessor individual differences on rating errors and accuracy was evaluated in a field study of assessment centers. Rater errors were based on archival data containing over 20,000 assessments of candidates for assembly line positions. Cronbach's four components of rater accuracy were determined by providing assessors with behavioral performance data for candidates who had previously performed in an assessment center. ^ Rater errors are assumed to be relatively stable rating characteristics. Rater errors investigated were halo error, range restriction, and leniency. Individual differences in ability and motivation were expected to predict rating biases. An interaction model combining assessor ability and motivation was utilized to predict rating accuracy. ^ Eighty-two assessors were measured on five scales: attributional complexity, conscientiousness, intelligence, need to evaluate, and self-monitoring. Assessors were provided with dimensional ratings for 12 candidates. Ratings were then compared to known true scores to determine accuracy. ^ It was hypothesized that assessors with higher levels of ability and motivation would make fewer rating errors. Regression analyses indicated attributionally complex assessors showed less range restriction. More intelligent assessors gave more lenient ratings. Halo error decreased as assessors' experience increased. ^ For rating accuracy, it was hypothesized that higher levels of ability and motivation would interact to predict greater rating accuracy. Moderate support was found for the ability x motivation interactions in predicting accuracy. Regression analyses revealed a mixed pattern of significant interactions. Need to evaluate seems to play an important role in rating accuracy and appeared in four of the five significant interactions. For less attributionally complex assessors with a higher need to evaluate, greater elevation, differential, and stereotype accuracy were found. More intelligent assessors with higher levels of need to evaluate showed an increase in differential accuracy. ^ Further studies are required to better understand the role of need to evaluate in rating accuracy. Results also suggest that the design of the assessment center and training of assessors may be key elements in producing accurate, unbiased ratings. ^
Davis, Mary Elizabeth, "Influence of assessor individual differences on rating errors and rating accuracy in assessment centers" (1999). ETD collection for University of Nebraska - Lincoln. AAI9952674.