Data Mining in the Social Sciences: An Example from Political Science

Authors: Anthony Scime (SUNY Brockport, Department of Computer Science), Gregg R. Murray (SUNY Brockport, Department of Political Science)

Abstract

This paper proposes and demonstrates data mining as an analytical tool for the social sciences. We use an iterative domain expert and data mining process to identify the variables in the data set important to answering the question at ha nd. We demonstrate the data mining of social science data with an example from political science using the American National Election Study (ANES), a rich but disparate source of information about Americans' vote choices. Specifically, we use data mining classification to construct a decision tree to select important predictors of the vote from the more than 900 items that compose the ANES. We conclude that data mining may be useful for more fully exploiting important data collected, but often unexploited, for the purpose of answering important questions in the social sciences.