Predicting Election Turnout

The goal is to devise an objective statistical test of a generating hypothesis-in this case, a latent decision to vote in an election. The hypothesis is used to generate a set of predictive items and to arrange them in the rank order of relevance. Factor analysis provides a first test of the a priori ranking in terms of the factor loadings of the first component. A second test of the a priori ranking is provided by the observed power of the items to predict voting turnout. These rankings are statistically summarized by coefficients of rank correlation, and, in the example used here, the significance test sustain the generating hypothesis

A priori and a posteriori · Econometrics · Factor analysis · Political science · Rank (graph theory) · Rank correlation · Ranking (information retrieval) · Relevance (law) · Set (abstract data type) · Statistical hypothesis testing · Statistics · Test (biology) · Turnout · Voting · Artificial Intelligence · Computer Science · Electoral Systems and Political Participation · Game Theory and Voting Systems · Mathematics · Psychology