ABSTRACT
The lack of a simple, direct link between stimulus and response in human evaluation of flavor has frustrated many product developers and marketing managers who want to predict customer purchase from ingredient formulae. Psychophysical models have been proposed to explain how stimulus energy is transformed to a subjective experience and how that experience is then translated into an observed response. It is also known that humans react differently to the same stimulus under different circumstances. For example, response tasks such as evaluating flavor recognition, flavor intensity, and flavor appreciation are dependent on stimulus context. Poulton [1] has shown that context is responsible for many biases in judging sensory magnitude. Lawless et al. [2-4] have asked whether context effects can be eliminated or minimized by training. One could imagine that sensory scientists involved in industrial product testing at least have come to grips with the problem since context effects pose a major obstacle to measuring what really matters for product development. Another hurdle in the acquisition of meaningful data for product development is the assumption of a direct correspondence between instrumental measurements and human evaluation. Fortunately there are occasions when physicochemical data have been modeled successfully to explain some sensory experience. On the other hand, there is still great frustration at having to admit that a parameter as measurable as flavor release in the MS-Nose (Micromass, Manchester, UK) experiment [5] does not necessarily correlate with flavor perception.
