| Researchers traditionally have had difficulty establishing statistical connections between attitudes and behavior, or even modeling the relationship. This study shows that attitudes are more complex than ordinarily accepted. It shows how a conation variable and techniques for dealing with "noisy" data help enable statistical connections.;The model is based on Fishbein and Ajzen's Theory of Reasoned Action. Attitudes have no direct effect on behavior; rather, attitudes have an indirect causal effect on behavior through behavioral intention, the endogeneous conation variable.;The data are from a survey of office automation at Stanford University, an environment appropriate to Fishbein and Ajzen's rational choice assumptions. Major factors are Beliefs, Affective Attitudes, Prior Experience, Organizational Norms, Behavioral Intentions, and Behavior. The key hypothesis is confirmed: there is no direct connection between Attitude and Behavior. The complex attitude formulation indirectly influences Behavior through Behavioral Intention.;Regression analysis is used for preliminary exploration and model formulation. The overall fits and R;The study demonstrates that LISREL holds promise for studying social and educational processes, especially where measurement error is high or multiple indicators are available to represent important conceptual variables that cannot be measured directly. It finds the Fishbein model appropriate for modeling automation behavior in a higher education setting, and it provides a basis for educational researchers to think about the role of attitudes, their formulation, and their relationship to behavioral outcomes. Based on the theoretical factors, the study discusses interventions for the practitioner contemplating resource allocations. |
