John A. Howard

Nonlinear Relations in a Complex Model of Buyer Behavior

Econometric techniques are used to test in a nonlinear fashion a model built around the learning subsystem of the Howard-Sheth theory of buyer behavior. Findings of nonlinearities (including interactions) are shown to improve the R 2, s , to be generally consistent with the theory, and to provide researchers with hypotheses to be tested in other purchase situations.

Some Empirical Contributions to Buyer Behavior Theory

A comprehensive buyer behavior model is examined in the context of a test market study utilizing a specially designed panel. The analysis uses both full-model cross-sectional and longitudinal procedures. Improvements for the hypothesized model are suggested, as well as implications for buyer behavior theories in general.

The Relationship of Liking and Choice to Attributes of an Alternative and Their Saliency.

Evaluations of specific attributes of a subcompact automobile were combined in linear regressions to predict overall liking and purchase intention. Of two forms-raw scales and scales weighted by the importance attached to each attribute by each subject-unweighted evaluations proved more consistent and important predictors than did these same evaluations weighted by their saliency.