Abstract
Summative, averaging, and elimination by aspects models of choice were applied to voter decisionmaking in the 1972 New Hampshire presidential primary. Predictions were generated from each of the models and compared to election outcome, candidate preference, and reported vote. The summative and elimination by aspects procedures predicted equally well, and both predicted significantly better than the averaging procedure. Candidate personal attributes were identified by the majority of voters as more important dimensions in deciding for whom to vote than candidate issue positions. Primary elections have many interesting features. Turnout is lower than in general elections. The voters are more involved, and tend to differ from participants in general elections in the same ways that general election voters differ from nonvoters (Ranney and Epstein, 1966; Ranney, 1968). In some states, participation is limited to partisans; in others, independents or members of the rival party may participate if they wish. Often more than two candidates compete for a party's nomination, and voters choosing between them lack the familiar guidepost of party identification. Even nonpartisan elections differ from primaries in this last respect, since voters often know the
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call