An Application of Item Response Theory to Consumer Judgment

Abstract

A probabilistic approach based on multidimensional item response theory (IRT) is proposed to model consumer judgment. Compared to multidimensional scaling and factor analysis, item response theory requires less input from subjects but provides equal or more information. Therefore, IRT holds great promise in applying to the marketing context. The process underlying consumer evaluation is hypothesized as the comparison between the product value a brand provides and the expectations characterizing a consumer. That is, the more positive the comparison, the higher the probability that the consumer has a favorable attitude toward the brand.

Use of item response theory in marketing research

“…Some examples of these papers, in the order of increasing complexity of the item response theory models, are as follows. Lou (1995) presented a brief review paper on item response theory for consumer judgment. Raykov and Calantone (2014) provided a more extensive review paper that discussed the utility of item response theory modeling in marketing research and presented many basic concepts and models within item response theory (see also Bacon & Lenk, 2008;Salzberger & Koller, 2013;Singh, 2004).…”