Probabilistic scaling analyses of sensory profile, instrumental and hedonic data

Sensory and hedonic variability are fundamental psychological characteristics that must be explicitly modeled if one is to develop meaningful statistical models of sensory phenomena. Sensory objects are perceived with differential uncertainty. Subjects also differ; some are very certain about what they desire and others are not. When variability is not modeled, the variability becomes confounded with dissimilarity or disutility and the estimates lose their meaning. Probabilistic scaling models are proposed that explicitly measure sensory and hedonic variability. The resulting estimates are shown to be free of the systematic bias that characterizes traditional deterministic models. Applications of the proposed models are presented and extensions of the models that can help bridge the gap between sensory science, marketing and product development are illustrated. Copyright © 2005 John Wiley & Sons, Ltd.