Abstract

Consumer perceptions and preferences are two critical constructs in marketing research because they are theorized to affect consumers' judgements, intentions to buy and actual purchase behavior. Recently, several researchers in multidimensional scaling (MDS) argued for the joint analysis of the perception and preference data collected from the same sample implicitly assuming that the stimulus spaces for perceptions and preferences were similar in terms of number of dimensions and coordinate values. Three studies are conducted to test this assumption. It is hypothesized that in MDS, stimulus spaces for perceived similarity data and preferences data would be congruent in terms of the number of dimensions and coordinate values for: (1) those stimuli that could be naturally described in terms of continuous attributes, and (2) for stimuli that are high in the choice hierarchy but (3) not for stimuli that are low in the choice hierarchy. The findings confirm the hypotheses regarding the first two types of stimuli but not the third one. Indeed, the stimulus spaces based on the perceptual and preference data for all three types of stimuli were similar in terms of the number of dimensions, estimated coordinate values and the meaning of the dimensions. This finding encourages further research regarding the development of MDS models jointly using a variety of different measures involving perceptions of similarity, ratings of attributes, and preferences. Limitations of the study, its implications and future research directions are discussed.