Potential Fallacy of Correlating Hedonic Responses with Physical and Chemical Measurements

Abstract

ABSTRACTPerceived intensity (strength) and hedonic responses (like/dislike) were obtained from 29 judges for three sensory systems: color in orange juice, oral viscosity of apricot nectar, and sweetness in lemonade. Intensity functions were directly related to concentration of the additives in each system, and were positively correlated (p < 0.001) with corresponding instrumental analysis—Hunter a values for orange juice, Brookfield viscosity at 30 rpm for apricot nectar, and refractive index of lemonade. Intensity judgments were stable for individuals and for the group from the beginning to the end of the 7‐wk study. In contrast, hedonic responses varied extensively among judges, showed changes in direction for some judges from the beginning to the end of the study, and the group means showed an inverse or no linear correlation with the corresponding instrumental values. Intensity and hedonic responses are distinct behaviors which exhibit different functions with stimulus concentration, usually linear for the former and cubic for the latter. In most sensory‐instrumental comparisons, linear correlation can be applied validly only to perceived intensity, within the linear portion of the usually sigmoidal curve but not to cubic or quadratic preference, acceptance, or hedonic responses.