FUZMAR: An approach to aggregating market research data based on fuzzy reasoning

The authors introduce an approach for validating models which contain variables taking linguistic values drawn from a scale which is a linear ordering. This approach uses operations from fuzzy logic to construct the models and the principal of minimal entropy to choose between competing models. The authors suggest the usefulness of this approach for market research surveys. They implement this approach in a test case using economic attitudinal variables to predict the purchase of major consumer products.