Key points are not available for this paper at this time.
Ideology is a central concept in research on American public opinion, yet it continues to be used in crude ways. Recognizing both the widespread use of the seven-point scale, and its numerous problems, we provide new findings on how ideology is used by survey respondents, the attitudinal correlates of ideological self-assignment, and the impact of education on ideological constraint. From these findings, we propose an alternative coding strategy that uses dummy variables for each ideological pole, which are then interacted with a three-point educational scale. Replicating past research on vote choice, racial attitudes, and respect for gay rights, we show that this new scheme generates more precise statistical estimates and offers new substantive insights into the relationship between ideology and a host of political outcomes.
Wood et al. (Sun,) studied this question.