Search

An experiment on taste sensory analysis of an artificially arranged compound of salty and sour flavors was conducted with the goal of evaluating sensory performance to a mixture of flavors akin to natural food, but that allows for keeping a higher experimental control than natural foods do. Participants performed a detection task in which the concentration of salt was manipulated between-subjects (from 0.07 to 0.75%). A procedure based on a dissociative model that allows for obtaining independent measures of sensitivity and criterion performance was used. In general, sensitivity index increased with the increase in signal salience, with actual performance approaching a Weibull distribution. A slight but reliable lenient response bias was found regardless of the salt concentration. The relevance of using independent measures of sensitivity and bias in sensory evaluation of complex compounds, and its relevance for sensory analysis of food is discussed. Practical Applications In tasting situations, perception seems to be easily influenced by cognitive factors and different response biases might appear, modifying the response criterion location. Difficulty of discrimination may also increases by qualitative changes that occur in taste mixtures. Signal Detection Theory allows the measurement of both: sensory and nonsensory components. Taking apart cognitive and sensory components is central to uncovering small differences in sensory qualities and developing accurate models of perception. This approach is of special interest to study the psychological processes involved in tasting situations, including the evaluation of training programs conducted to improve tasters' discriminative abilities. In addition, function modelization has important implications for the area of perceptual research since psychophysical functions for binary compounds as it allows conducting predictions about interaction effects between studied components in complex stimuli. [ABSTRACT FROM AUTHOR]

Human participants were trained in a trial-by-trial contingency judgements task in which they had to predict the probability of an outcome (diarrhoea) following different cues (food names) in different contexts (restaurants). Cue P was paired with the outcome on half of the trials (partial reinforcement), while cue C was paired with the outcome on all the trials (continuous reinforcement), both cues in Context A. Test was conducted in both Context A and a different but equally familiar context (B). Context change decreased judgements to C, but not to P (Experiment 1). This effect was found only in the cue trained in the context where a different cue was partially reinforced (Experiment 2). Context switch effects disappeared when different cues received partial reinforcement in both contexts of training (Experiment 3). The implications of these results for an explanation of context switch effects in terms of ambiguity in the meaning of the cues prompting attention to the context (e.g., Bouton, 1997) are discussed. [ABSTRACT FROM AUTHOR]