Robustness and Sufficiency of the Theory of Reasoned Action in Longitudinal Prediction

An examination of the theory of reasoned action (Fishbein & Ajzen, 1975) applied to longitudinal data is reported. Data consist of responses of a national sample of high school sophomores concerning their college plans and subsequent achievements. Subjects initially were queried in 1955 and followed up 15 years later. Measurement robustness of the theory constructs was evidenced by confirmation of the predicted relationships despite novel and crude measures. The predictive utility of the theory was demonstrated in substantial prediction of both college attendance (2-3 years subsequent to the measures) and the ultimate level of education attained in the following 15 years. The sufficiency claims of the theory were challenged by the finding of a direct subjective norm-behavior relationship unmediated by intention. The impact of two exogenous, demographic predictors was mediated by this expanded model.