Doctoral Dissertations

Date of Award

8-1989

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Industrial and Organizational Psychology

Major Professor

Michael C. Rush

Committee Members

H. Dudley Dewhirst, Robert T. Ladd, Eric Sundstrom

Abstract

The objective of this study was to provide a test of an information processing model for performance appraisal proposed by Feldman (1981). Feldman proposed that an observer cognitively categorizes an actor and then uses a category prototype as a cognitive representation of the actor during the appraisal process. Feldman's model suggests that an observer will use either an automatic or controlled process in assimilating information depending on the consistency of information. Feldman's model suggests that an observer will rely on an automatic information processing function if salient attributes of an actor correspond to the attributes associated with an existing cognitive category. However, an observer is expected to use a controlled information processing function to assign the actor to a cognitive category if characteristics of the actor do not easily conform to those of an existing category or the actor exhibits behavior that is inconsistent with the observer's category influenced expectations for the actor.

The present study examined the effects of information inconsistent with an actor's typical behavior on the categorization and evaluation process across time. The sample for this study consisted of 120 graduate teaching assistants at The University of Tennessee. The experimental design was constructed as a partially crossed factorial design consisting of two levels of induced initial attribution conditions (i.e., internal and external), initial performance (i.e., high and low), and instances of discrepant information (i.e., single and multiple) introduced at Time 2 and Time 3. A Repeated Measures Analysis of Variance was employed to test the results of the study against theoretical predictions.

It was proposed in the study, following Feldman (1981), that the participants would use causal attributions for an actor's performance as a source of information in initially assigning the actor to a cognitive category. The participants were also expected to use the characteristics associated with an actor's assigned cateory (i.e., person impression) when asked to evaluate the actor. It was also expected that sufficient discrepant information about the actor would elicit a change in initial attributions and evaluations (i.e., re-categorization of the actor). It was expected that multiple instances of discrepant information would elicit a change in attributions for participants in the initial external attribution condition at Time 1 to an internal attribution attribution at Time 3; while participants in the initial internal attribution condition would attribute the actor's performance to internal causes at Time 1 and Time 3. It was also expected that changes in the participants' attributions for the actor's performance from Time 1 to Time 3 would be associated with changes in the participants' evaluation of the actor.

Contrary to expectations, the results did not reveal that the participant's responses on the dependent measures varied across time as a function of initial attributions and amount of discrepant information. The results, instead, indicated that the responses of participants who received multiple instances of discrepant Information varied across time as a function of initial performance. Participants in the initial low performance condition who received multiple instances of discrepant information (1) reported that the actor's performance was more a function of internal causes at Time 3 than at Time 1 and (2) rated the actor higher on the person impression measure at Time 3 compared with Time 1 regardless of initial attributions. Similarly, participants in the initial high performance condition who received multiple instances of discrepant information (1) reported that the actor's performance was more a function of external causes at Time 3 than at Time 1 and (2) rated the actor lower on the person impression measure at Time 3 compared with Time 1. The responses of participants in the single instance of discrepant information condition regardless of initial attributions and initial performance level remained relatively constant from Time 1 to Time 3.

The results of the study seemed to indicate that an observer may use causal attributions as a source of information when initially assigning an actor to a cognitive category. However, the results also suggest that the actor's initial and subsequent level of performance may contribute more strongly to the observer's categorization of the actor. The results are discussed in terms of the suggested need for additional research to investigate the relationship of cognitive information processing and the performance appraisal process as a function of encountered discrepant information.

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

Share

COinS