A place for emotion in attitude models

Finding ways and means to incorporate emotional experience into consumer and market research has been an ongoing challenge. We frame this challenge in the context of the information that is and is not normally collected in multiattribute attitude models. Our data show that retrospective reports about emotional experiences can be useful predictors of attitude when compared to traditional measures of cognitive structure, and that prior experience with a behavior can play a major role in moderating these relationships. The appeal of multiattribute attitude models has always been their value for predicting and diagnosing motives and preferences. Integrating emotive information into these models appears to be a way to build on this appeal. (c) 2003 Elsevier Inc. All rights reserved.