Abstract
A multiplicative model is proposed for the similarity between stimuli varying in easily discerned dimensions. According to this model, the multidimensional similarity estimate equals a product, weighted by exponents, of the unidimensional estimates. The multiplicative model, together with two alternative ones, a vector model and an additive model, are tested in two experiments where the stimuli are varied in two dimensions. The multiplicative model seems to describe the empirical data most closely.
Get full access to this article
View all access options for this article.
