A systematic evaluation of the components of frame-of-reference training and their effects on rating error, accuracy, and individual cognitive processes.
Item
-
Title
-
A systematic evaluation of the components of frame-of-reference training and their effects on rating error, accuracy, and individual cognitive processes.
-
Identifier
-
AAI9130321
-
identifier
-
9130321
-
Creator
-
Hartog, Sandra B.
-
Contributor
-
Adviser: Roger E. Millsap
-
Date
-
1991
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Psychology, Industrial
-
Abstract
-
Research in the performance appraisal field has been moving away from issues that are most salient for organizational use and toward a more theoretical investigation of underlying processes in performance ratings. In this study, Frame-of-Reference training (FOR) was investigated as a technique that can be directly applied to increasing the effectiveness of organizations' performance appraisal systems. FOR training was chosen because its methodology exemplifies many of the separate elements employed in performance appraisal training programs, and thus the results may be generalizable to other research in this area. This study was also undertaken to clarify many inconsistencies in this research paradigm, including the generalizability of effectiveness indices, operational and conceptual ambiguities across training programs, and elements in the research design itself (e.g., topic and order effects of videotaped performance vignettes).;The components of FOR training, as originally conceptualized by Bernardin and Buckley (1981), were systematically investigated to determine the differential influence of each on error and accuracy effectiveness indices. Trainees were 250 college students across 12 intact classes. Two classes were randomly selected for each of six training conditions. The difference between conditions was the omission of one training component each. The effectiveness indices were measures of leniency, interrater reliability, two indices of halo error, elevation, differential elevation, differential stereotype, distance and leniency accuracies, and a measure of absolute value halo.;The results indicate that the relationship between the effectiveness of the various components involves an interplay between the types of rating judgements required, the integration of similar or different types of information, and the underlying cognitive strategies promoted by each component. When FOR training fares best it appears due to the feedback of expert true scores and behavioral rationales, along with participant discussion. When FOR training leads to inaccuracy, different combinations of components differentially influence the accuracy indices. The results also suggest that the choice and presentation of stimulus materials are very important design considerations. Additional results demonstrate that: rater attitudes are related to accuracy and dimension ratings; there are differential relationships between identical indices of observation and evaluation accuracies; and, the different evaluation accuracy indices do not generalize across measures.;Contributions of this project are discussed, as are limitations. Numerous suggestions for future research are offered.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
