THE EFFECTS OF MANIPULATING ATTRIBUTIONS UPON SUBSEQUENT SOCIAL PERFORMANCE

BRAD KEITH GRUNERT, Marquette University

Abstract

The application of attribution theory to skilled social performance is a logical area of investigation in light of the emphasis of recent research in each of these areas. Social interactions have been conceptualized as being a form of an achievement situation and, as such, should result in the formation of attributes similar to those present in other achievement settings. Little research into this area has been conducted. The purpose of this study was to examine the effects of manipulating the attributional components of motivation upon socially skilled performance. The subjects were 180 male undergraduates of Marquette University between the ages of 18 and 25. Each subject participated in a 5-minute interaction with a confederate who was instructed to respond to the initiations of each subject but not to initiate any conversation herself. Following a 10-minute waiting period, during which subjects were told a panel of judges was rating their interactions, one of nine arbitrary feedback conditions was administered. The subjects then engaged in a second 5-minute interaction with the same confederate. When this was finished, the subjects were asked to complete a questionnaire in which they rated how each of four attributes accounted for their performance. The independent variable for this study was the type of feedback the subject received (positive or negative ability, effort, task difficulty, or luck or the neutral control). The dependent variables were the number of initiations for each interaction and the attribution ratings. The data for the number of initiations for each interaction were analyzed by means of a 2 x 2 x 2 factorial analysis of variance with repeated measures and nested subjects. The results of this yielded significant within subjects, subjects by valence, and subjects by valence by stability effects (p < .05). The data from the attribution ratings were analyzed by means of a multivariate profile analysis. Significant profile differences resulted from the locus and valence factors (p < .05). The results obtained were consistent with the predictions of attribution theory. The significant findings were discussed and their implications for attribution theory were investigated. Suggestions for future research in this area were outlined.

This paper has been withdrawn.