AN INDUCTIVE APPROACH TO DETERMINING CAUSALITY IN COMPARATIVE POLITICS: A FUZZY SET ALTERNATIVE

摘要

Political science typically tests hypotheses by demonstrating correlations between variables. The most commonly employed method for doing so is regression analysis. The approach is essentially crisp, which carries with it a number of questionable assumptions about the data. Political phenomena such as democracy or stability often involve measuring the degree to which a particular observation demonstrates the qualities of the category. A fuzzy set approach better captures the inherent ambiguity in classifying our observations relative to such categories. However, the attempt to establish correlations between fuzzy sets in the social sciences has been plagued by the priority ranking issue. We illustrate the potential that Jeffrey's Rule has to overcome this difficulty.