Open Collections

An application of Fishbein's attitude theory to the prediction of free-choice student behaviors in a first year university physics course

University of British Columbia

The purpose of the study was to forecast the actual performance of five extracurricular educational activities by 128 first year university Physics students using Fishbein's model for the prediction of behavior and behavioral intention. The effectiveness of achievement measures and measures of attitude toward various instructional objects in the prediction of behavior and behavioral intention was also investigated. Consideration of Fishbein's model led to the investigation of several specific problems: (a) the relationship between variables internal to and those external to the model; (b) the relationship between behavior, behavioral intention, and the attitudinal and normative variables of the model; (c) the accuracy with which behavioral intention and behavior could be predicted, and the relative importance of the predictors in the prediction equation; (d) the use of behavioral intention measures as predictors of behavior in specific educational situations; and (e) the detection of possible measurement effects. A Likert attitude scale was used to obtain measures of attitude toward fourteen different aspects of Physics and Physics instruction. Estimates of Grade 12 Mathematics and Grade 12 Physics achievement were obtained from self- reports. Fishbein's model was applied to measures of: students' attitudes toward performing each activity (A act), their social normative beliefs (NB s), personal normative beliefs (NBp),motivation to comply with certain referents (Mc ), and behavioral intention (BI). Behavioral intentions were also predicted for three of the voluntary activities, using measures of A act, NBs and NBp as predictor variables. The measures of normative beliefs were taken with respect to the referents: self, closest friends, parents, majority of the class, lecturer, and religious group. The model for predicting behavioral intention was given by Fishbein in the form of a multiple regression equation, where the criterion variable is BI and the predictor variables are Aact and the summation (over all referents) of NBs multiplied by Mc. Most of the obtained results tended to agree with expectations based on Fishbein's theory. Variables external to the model were, for the most part, poorly correlated with behavioral intention and with overt behavior (B) unless they were significantly correlated with at least one of the predictors given in the model. Statistically significant correlations were consistently found between measures of BI and NBp, Aact, and the normative belief with respect to students' 'best friends'. The magnitudes of correlations between measures of BI and the other social normative beliefs varied considerably across activities, several correlations reaching statistical significance. Correlations between B and measures of BI were generally low, although three out of five were significantly greater than zero. Correlations between behavior and the predictor variables were also small, and were frequently not statistically significant. High multiple correlations obtained in the prediction of BI indicated predictive validity of the predictor variables. In all predictions of BI, NBp had, by far, the greatest weight as a predictor. Beta weights of Aact, and NBs varied greatly across activities. Low multiple correlations were obtained in the prediction of behavior from the predictor variables, substantiating the low product moment correlations obtained between BI and B. The observation that significant positive correlations between behavior and the predictor variables were reduced to nonsignificance when behavioral intention was held constant, tended to substantiate the theoretical expectation that BI is an intervening variable between behavior and the predictor variables. An unexpected result was the detection of significant measurement effects in the prediction of voluntary performance of three activities. These effects were substantiated by means of χ²tests of the independence of behavioral responses obtained under different measurement conditions: administration of the research instrument, a placebo instrument, and no instrument. It was concluded that with the application of Fishbein's theory, the prediction of behavioral intention with respect to performing free-choice activities in an educational setting could be made with considerably better than chance accuracy. The prediction of actual performance of the activities from measures of behavioral intention, however, posed serious difficulties. It was recommended that the possibility of measurement effects influencing the prediction of behavior be given careful consideration in future educational applications of the model.

Text

2011-04-06

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/33339

Education

University of British Columbia

Graduate