The Statistical Analysis of a Disparity Test

Abstract

A market research survey was conducted using a Disparity test to measure the change in preference for four consumer products, as the products became progressively "worn" through use. The results were analysed to see if the deterioration in average preference for a product was uniformly less than for another product, in order to see if "longer lasting" claims for this product could be made for one of the dimensions measured in the test. By fitting a suitable statistical model to the data using weighted least squares estimation it was possible to infer that it was reasonable to make such a claim for one of the products.