| Abstract: | Measures of change in pairs of attitudinal variables can provideimportant insights into the structure of the political beliefsystems of mass publics. Panel data reveal evidence of the greatercentrality of some idea elements rather than others in the contextof short-term dynamic constraint. Specification of the theoreticallyrelevant voter attributes makes it possible to test for expectedstructural differences connecting policy related predispositionsand policy preferences; specification also makes it possibleto test propositions involving the reciprocal effects of attitudesand emerging vote preferences. Some of the more helpful specificationsdisclose the extent to which population heterogeneity producesa blurred image of relationships when analysis is based on thetotal electorate rather than limited to voters or subsets ofvoters specified by theoretical criteria. |
