Modelling product positions through the use of multidimensional scaling techniques: An empirical investigation

Abstract

This research is concerned with two approaches to modelling the market for king size cigarettes in the UK. In the first approach overall perceptions of the similarity between brands are analysed by multidimensional scaling techniques (using the MDSCAL V computer package) to develop a perceptual map of the market under investigation. The dimensions of this perceptual map are then suggested by external property fitting (using PREFMAP PHASE IV). In the second approach perceptions of the brands in the market are considered on each of a set of seventeen bipolar product attributes generated from previous, qualitative research. These perceptions are then used to model the market and to describe the dimensions by an internal property fitting analysis (using MDPREF). Results from the two approaches are compared across two independent samples, each broadly representative of the market in product usage terms, for an indication of internal validity of the models.