The voter's guide to the galaxy—a multiple-criteria fuzzy decision-support tool for voters and a fresh take on election survey methods

Abstract

This paper suggests a multiple-criteria decisionsupport tool for voters, that compares the attitudes of the voters with the declared attitudes of the political parties in several sets of relevant issues. The model intends to identify parties that seem to provide the best fit with the voter attitude-wise. The data input methodology uses discrete 5-point Likert-type scales. We investigate the effect of the inclusion of weights of different sets of issues, of the numerical anchors of the values of the Likert-type scales and also of the potential presence of extremity/leniency effect on the suggestion of the "most compatible" political party suggestion. We also propose a simple fuzzyrule based evaluation tool to identify serious incompatibilities or desired compatibilities in the attitudes of the voter and the party to the relevant issues. This tool introduces (un)acceptability thresholds for the differences in attitudes between the parties and the respondents and provides lists of parties to vote for or to avoid voting for accompanied by the strengths of these suggestions. The tool is shown to have several desirable features including lower sensitivity to small differences in the attitudes, respondents' ability to express their preferences and also preventing the compensation of unacceptable differences in some categories of important issues by high compatibility in the other categories.