Consistency of Consistency

This project explores the topic of what causes respondents in a survey situation to answer more or less consistently across time to the identical political issue questions. This discussion was begun by Philip Converse in 1964, who used data from the 1956-1958-1960 American National Election Studies panel survey to put forth the idea that certain respondents were generally more consistent across time than others, and that these more consistent respondents should be characterized by a higher level of political attunement, which would involve such factors as education and interest in politics. Utilizing the same data, this project uses regression analysis to discover significant correlations between respondents’ temporal response consistency and their demographic and other characteristics. Results suggest that instead of being generally related to education and political interest, temporal response consistency depends on the relevance of the particular issue at hand to the respondent’s life and the clarity of the positive or negative consequences of the issue for the respondent and the respondent’s community.