Master's Theses

Document Type

Thesis - campus only access

Date of Award

Summer 1985

Degree Name

Master of Science (MS)

Department

Psychology

Advisor

Stephen Klein

Abstract

The purpose of the present study was to replicate and expand on a study conducted by Feltz (1980). A path analytic procedure was utilized to evaluate the effectiveness of three theoretical models (an anxiety-based, Bandura, and Feltz) in predicting approach/avoidance behavior with 40 male and female highly test-anxious college students exposed to an anagram solution task. An anxiety model, based on Mowrer’s two-factor theory and a “three-systems” approach to anxiety, assumed that cognitive anxiety, physiological arousal, and previous related performance all have a direct causal effect on avoidance behavior. Self-efficacy was only considered an effect of these three anxiety components. For the Bandura model, previous performance and automatic arousal indirectly influenced avoidance performance via self-efficacy. Self-efficacy was presumed to be the only variable to have a direct effect on avoidance performance. The Feltz model, which was a respecification of Bandura’s self-efficacy model, assumed that previous anagram performance had both a direct and indirect influence by way of self-efficacy on succeeding performance. Physiological arousal was hypothesized to predict the initial anagram attempt only. The results did not totally support the three models. Cognitive anxiety and physiological arousal were not meaningful direct predictors of avoidance behavior. Self-efficacy directly influenced the initial anagram performance but it had little effect on later performances. Self-efficacy was found to be more an effect of performance and cognitive anxiety after the first anagram performance. For all variables investigated, previous performance was the best direct predictor of subsequent anagram performance. The anxiety-based and Feltz models similarly accounted for the most variance in avoidance performance. However, the Feltz model provided a more parsimonious explanation for avoidance behavior, and it contained more variables that significantly influenced performance than the anxiety-based model.

Comments

For questions contact ScholarsRepository@fhsu.edu

Rights

© 1985 Ronald T. Berg

Download

Off Campus FHSU Users Click Here

Share

COinS