The joint spatial representation of multiple variable batteries collected in marketing research

Marketing research studies pertaining to market segmentation, competitive market structure, and product/service positioning often involve the collection of multiple batteries of measurements from the same set of respondents (e.g., preferences, attribute ratings, demographics, proximity judgments). Various multidimensional scaling (MDS) and spatial methods have been used in the analysis of these variable batteries separately, but not much effort has been expended in attempting to relate the individually derived spatial structures. The authors propose a new latent structure MDS procedure that is devised to represent jointly the structure in multiple batteries (preferences, proximities, and brand attribute ratings) of variables collected across the same set of respondents. The authors present the technical structure of the proposed maximum likelihood-based model and conceptually compare it with other related spatial MDS models. The authors present an illustration of the procedure with respect to the analysis of published proximity, preference, and brand ratings data collected from consumers who evaluated ten brands of soft drinks.