The k-Visit Method of Consumer Testing

Publisher Website
Google Scholar
Abstract
When testing a pair of products for consumer preference the problem of how to treat or interpret no preference votes arises. A method is described of collecting data from a given number of consumers by repeated visits to them, or by obtaining repeated judgments from them in stores, which enables the estimation of the true proportion of those consumers in the population who have a preference for either product and of those who cannot discriminate or have no preference. For the model assumed the maximum likelihood estimators of the above proportions are derived, their variance-covariance matrix is obtained, and a way of testing the appropriateness of the model is indicated. A decision theoretical formulation is suggested.

This publication has 0 references indexed in Scilit: