Key points are not available for this paper at this time.
Attitudes can be seen as object evaluations stored in memory. Accordingly, attitude structure may be seen as a memory structure with dynamic implications for information processing. In this article, an associative network model for the representation of stored attitudes was assumed, and 2 manifestations of the notion of spreading activation within the confines of such a model were examined. Study 1 demonstrated that giving one attitude response could facilitate a 2nd attitude response if the 2 shared a structural link in long-term memory. Studies 2 and 3 showed that spreading activation could also result in polarization o fattitude responses, so that answering questions on 1 attitude issue might result in more extreme responses to a 2nd linked attitude issue. These results have implications not only for theories about attitudes but also for measurement issues in political survey research. For many years, social psychologists have been interested in the structure of attitudes. In the literature that has accumulated on this topic, attitude structure is discussed in two quite different ways. Some researchers focus on the internal structure of a single attitude, viewing it as a composite of beliefs about the attitude object, affective responses to the object, and behavioral tendencies (e.g., Breckler, 1984; Katz Cartwright Festinger, 1957; Heider, 1958). The latter approach to attitude structure, on which we focus in this article, has been of substantial interest outside psychology. For example, it is often cited in an important debate within political science concerning the degree to which the general public's attitudes on government policy issues are derived from or structured by broad ideological principles such as liberalism and conservatism.
Judd et al. (Fri,) studied this question.