Date of Award

Fall 1985

Project Type

Dissertation

Program or Major

Psychology

Degree Name

Doctor of Philosophy

Abstract

Two experiments were conducted to test the predictive validity of a model put forth by Alloy and Tabachnik (1984) to account for judgments of contingency. The authors contend that such judgments can be accounted for by determining the interaction of information contributed by "one's expectations" and by the objective "situation." It was argued here that those two components, independently or interactively, are not always sufficient to predict judgments; the same objective data may be judged differently in different contexts and sometimes contrarily to one's expectations. That is, contrast effects may occur. This argument was supported using two different paradigms. In the first experiment, 66 female Introductory Psychology students were evenly divided into three groups which each received two sets of 40 trials. In the first set of trials, the response-outcome contingency was high, low, or absent. In the second set, the contingency between response and outcome was absent. The results revealed that subjects who initially received high-contingent response-outcome trials judged Set 2 non-contingent response-outcome trials as having significantly less contingency than subjects who initially received non-contingent response-outcome trials. The second experiment conceptually replicated and extended those findings. Thirty female Introductory and Social Psychology students were evenly divided into two groups who initially received exposure to 40 trials in which response-outcome contingency was high or absent followed by 40 more trials in which response-outcome contingency was moderate. The Set 2 judgments provided by each group were significantly different in that subjects with initial high-contingent experience judged later medium-contingent response-outcome trials as having significantly less response-outcome contingency than did those participants who had initial exposure to non-contingent response-outcome trials. Alloy and Tabachnik's model apparently cannot account for these findings. A comprehensive review of the literature is included as well as implications for that model and directions for future research.

Download

Share

COinS