DIFFICULTIES AND A WAY OUT: A BAYESIAN APPROACH FOR SENSORY DIFFERENCE AND PREFERENCE TESTS

There are some theoretical difficulties in the philosophy and methodology of sensory difference and preference tests. The difficulties lie in the assumption that the subjects have the same response ability, which is in conflict with psychological ideas about the processes of perception and decision. A Bayesian approach may overcome these difficulties by treating the parameter of proportion as a random variable. This paper presents a Bayesian approach for analysis of data from such sensory tests.     

SENSORY ANALYSIS IN MARKETING PRACTICE: COMPARISON AND INTEGRATION

Sensory analysis has traditionally played a prominent role in quality control for food products. More and more principles from sensory analysis are also applied in the area of food product development, bringing sensory analysis more closely into the domain of marketing. Unfortunately, in practice integration between sensory and marketing practices is far from optimal. Differences in basic orientations between sensory and marketing are a major source contributing to this defective cooperation.
Sensory analysis has traditionally been product oriented with an emphasis on internal validity of the test results. Implicitly or explicitly this approach emphasizes the relationships between sensory evaluation and characteristics of the product. Marketing, on the other hand, stresses the external validity of test results: the extent to which test results can be generalized to market behavior. Emphasis on external validity requires an approach to sensory analysis that is fundamentally different from current sensory practice in terms of type of respondents, type of stimuli, scaling procedures adopted and test circumstances.
Closer integration between the product and consumer oriented approach to sensory analysis may contribute to the success of product development in the food industry. The literature suggests several factors that may contribute to a more fruitful cooperation between the two approaches to sensory analysis. The company's senior management plays a central role in the achievement of this integration by providing an infrastructure (in terms of personnel, organizational structure and financial resources) that paves the way for closer cooperation.     

SENSORY ADAPTATION AND THE STATIONARY STATE

1. Experiments are described which measure the sensitivity of animals exposed to continued illumination to which they have become adapted. It is shown that the amount of outside light energy necessary to stimulate an adapted animal increases with the intensity of the adapting illumination. 2. The data are analyzed quantitatively in terms of the reversible reaction S ⇌ P + A shown previously to account for the photic sensitivity of these animals. This analysis demonstrates that, though the amount of incident energy necessary for a minimal response varies with the adapting intensity, the actual amount of photochemical decomposition required to set off the sensory mechanism is a constant quantity. 3. The ability of these animals to come into sensory equilibrium with any sustained illumination is accounted for quantitatively by the presence of a stationary state in the reversible photochemical reaction S ⇌ P + A during which the concentrations of the three components are constant. 4. It is shown that the concentrations of these substances at the stationary state are automatically controlled by the outside intensity. Therefore, given the sensory mechanism as a basis, the adaptation of the animals to light and the consequent changes in sensitivity, are determined entirely by the light to which the animals are exposed. 5. Because of the properties of the stationary state, and of the constancy of photochemical decomposition for a minimal effect, it is suggested that the sensory system is not only the traditional receptor system, but is also a protecting layer which stabilizes and buffers the relation between the nervous system and the environment.     

SENSORY ADAPTATION

Sensory evaluation makes use of the remarkable virtuosity and range of the human senses as a multi-purpose instrument for measuring the sensory characteristics of foods. The brain protects itself from an overload of information from the senses by two processes: feature extraction and adaptation. The former involves information reduction by the extraction of selected features from the environment; these form the basis for the reconstruction of the percept in consciousness. The latter, adaptation, involves the attenuation of repetitive and constant input so as not to overload the brain with redundant information.
The effects of adaptation can be observed for all senses. For the chemical senses, the effect is that a constant odor or taste stimulus will be perceived as decreasing in intensity while sensitivity to that stimulus is also decreased. For sensory evaluation, this poses problems. It means that a taste or odor has a tendency to vanish while it is being observed and that sensitivity to subsequent stimuli will be altered. Such sensitivity drift in the human instrument must be anticipated in the design of measurement procedures for the sensory evaluation of food.
For taste, adaptation changes caused by the measurement method can be seen to be largely responsible for disagreements in the literature concerning threshold and intensity measurement. Adaptation is also a contributing factor, but by no means the only one, in determining the relative discriminability of sensory difference tests. It is worth noting, however, that adaptation is not always a disadvantage; it can sometimes be used to advantage in sensory testing procedures.     

SELECTION AND SCREENING OF A DESCRIPTIVE PANEL FOR EWES MILK CHEESE SENSORY PROFILING

The aim of this work was to select a competent sensory panel considering its ability to describe ewes milk cheese sensory properties. The panelists evaluated the sensory characteristics of several ewes milk cheeses in terms of odor, flavor and texture. The performance of judges was assessed taking into account the following aspects: (1) individual discriminatory ability and reproducibility for each sensory attribute; (2) group discriminatory ability and reproducibility and agreement between judges for each attribute; (3) panel homogeneity; (4) panel consonance; and (5) overall panel discriminatory ability. The panel performance was checked with a set of statistical procedures. The results obtained by the selected group were better than the initial panel from the point of view of discriminant power, reproducibility and agreement. The suggested method can be easily applied in routine control of assessors performance and to discover deviant behavior or differences in individual perception of the samples.     

SENSORY CHARACTERISTICS AND CONSUMER LIKING FOR CEREAL BAR SNACK FOODS

This study examined sensory aspects of cereal bar snacks for their relative importance to consumer perception and degree of like (DOL). Principal components analysis of conventional profiling data showed a distinct location of each bar type in multivariate space, separation being based on textural aspects for the hard crunchy bar and flavor dimensions (cereal, nutty versus chocolate. sweet, fruit) for others. A consumer panel ( n =56) examined a subset of the bars according to a balanced incomplete block design. Analysis of variance showed that the chewy, nutty and chocolate bars were liked most ( p <0.01) and this was linked to ideal levels of filling flavor and quantity, chewy and crunchy textures and sweetness. The majority of consumers ranked 'taste' as the most important characteristic influencing their purchase intent, followed by textural features, 'price' and 'appearance' . The 'healthy image' aspect was relatively less important. Analysis of the relationship between the sensory measures and DOL, using partial least squares regression, confirmed the above segregation of preference, with sensory aromas and flavors having most influence on consumer liking.     

A THURSTONE-COOMBS MODEL OF CONCURRENT RATINGS WITH SENSORY AND LIKING DIMENSIONS

A popular product testing procedure is to obtain sensory intensity and liking ratings from the same consumers. Consumers are instructed to attend to the sensory attribute, such as sweetness, when generating their liking response. We propose a new model of this concurrent ratings task that conjoins a unidimensional Thurstonian model of the ratings on the sensory dimension with a probabilistic version of Coombs' (1964) unfolding model for the liking dimension. The model assumes that the sensory characteristic of the product has a normal distribution over consumers. An individual consumer selects a sensory rating by comparing the perceived value on the sensory dimension to a set of criteria that partitions the axis into intervals. Each value on the rating scale is associated with a unique interval. To rate liking, the consumer imagines an ideal product, then computes the discrepancy or distance between the product as perceived by the consumer and this imagined ideal. A set of criteria are constructed on this discrepancy dimension that partition the axis into intervals. Each interval is associated with a unique liking rating. The ideal product is assumed to have a univariate normal distribution over consumers on the sensory attribute evaluated. The model is shown to account for 94.2% of the variance in a set of sample data and to fit this data significantly better than a bivariate normal model of the data (concurrent ratings, Thurstonian scaling, Coombs' unfolding model, sensory and liking ratings).     

RELATIONSHIP BETWEEN SENSORY AND INSTRUMENTAL TEXTURE PROFILE ATTRIBUTES

Texture relationships were studied using both sensory and instrumental texture profile analysis (TPA) techniques to evaluate twenty-one food samples from a wide variety of foods. High linear correlations were found between sensory and instrumental TPA parameters for hardness (r = 0.76) and springiness (r = 0.83). No significant correlations were found between sensory and instrumental TPA parameters for cohesiveness and chewiness. Logarithmic transformations of data improved correlations between sensory attributes and their instrumental corollaries. The correlation between sensory hardness and the logarithm of instrumental hardness was improved to r=0.96. The correlation between the logarithm of both sensory and instrumental springiness was improved to r=0.86. The correlation between the logarithms of both sensory and instrumental chewiness was improved to r=0.54, which was significant at P<0.05.     

COMPARISON OF ATTRIBUTE LIKING AND JAR SCALES TO EVALUATE THE ADEQUACY OF SENSORY ATTRIBUTES OF MILK DESSERTS

Just-about-right (JAR) scales and attribute liking questions are usually used to study consumer perception of the sensory characteristics of food products. The aim of the present work was to compare the performance of attribute liking and JAR scales to evaluate consumers' perceived adequacy of flavor and texture of milk puddings. Two groups of consumers were asked to evaluate eight milk desserts using (1) overall liking followed by attribute liking for texture and flavor and (2) overall liking followed by JAR scales for thickness, creaminess, sweetness and vanilla flavor. Overall liking scores were significantly different when JAR scales or attribute liking questions were considered. Texture, flavor and overall liking scores were highly correlated to each other, providing the same information. JAR scales correlated better with the intensity of sensory attributes evaluated by a trained sensory panel, being JAR percentages a reliable tool to study the adequacy of sensory attributes.

PRACTICAL APPLICATIONS

Results from the present work showed that consumers might not be able to independently evaluate their liking of different sensory attributes of a product. For this reason, the use of attribute liking questions for studying the adequacy of sensory attributes in complex products would not be recommended. JAR scales were better indicators of the adequacy of sensory attributes. Consumers were able to independently evaluate texture and flavor attributes using JAR scales. However, the influence of JAR scales on overall liking scores should be taken into account when including these scales on consumer studies.     

A free energy principle for the brain.

By formulating Helmholtz's ideas about perception, in terms of modern-day theories, one arrives at a model of perceptual inference and learning that can explain a remarkable range of neurobiological facts: using constructs from statistical physics, the problems of inferring the causes of sensory input and learning the causal structure of their generation can be resolved using exactly the same principles. Furthermore, inference and learning can proceed in a biologically plausible fashion. The ensuing scheme rests on Empirical Bayes and hierarchical models of how sensory input is caused. The use of hierarchical models enables the brain to construct prior expectations in a dynamic and context-sensitive fashion. This scheme provides a principled way to understand many aspects of cortical organisation and responses. In this paper, we show these perceptual processes are just one aspect of emergent behaviours of systems that conform to a free energy principle. The free energy considered here measures the difference between the probability distribution of environmental quantities that act on the system and an arbitrary distribution encoded by its configuration. The system can minimise free energy by changing its configuration to affect the way it samples the environment or change the distribution it encodes. These changes correspond to action and perception respectively and lead to an adaptive exchange with the environment that is characteristic of biological systems. This treatment assumes that the system's state and structure encode an implicit and probabilistic model of the environment. We will look at the models entailed by the brain and how minimisation of its free energy can explain its dynamics and structure.     

APPLICATION OF FUZZY SETS AND NEURAL NETWORKS IN SENSORY ANALYSIS

This paper describes an initial but fundamental attempt to lay some groundwork for a fuzzy-set-based paradigm for sensory analysis and to demonstrate how fuzzy set and neural network techniques may lead to a natural way for sensory data interpretation. Sensory scales are described as fuzzy sets, sensory attributes as fuzzy variables, and sensory responses as sample membership grades. Multi-judge responses are formulated as a fuzzy membership vector or fuzzy histogram of response, which gives an overall panel response free of the unverifiable assumptions implied in conventional approaches. Neural networks are used to provide an effective tool for modeling and analysis of sensory responses in their naturally fuzzy and complex forms. A maximum method of defuzzification is proposed to give a crisp grade of the majority opinion. Two applications in meat quality evaluation are used to demonstrate the use of the paradigm and procedure. It is hoped that this work will bring up some new ideas and generate interest in research on application of fuzzy sets and neural networks in sensory analysis.     

Pollution going multimodal: the complex impact of the human-altered sensory environment on animal perception and performance

Anthropogenic sensory pollution is affecting ecosystems worldwide. Human actions generate acoustic noise, emanate artificial light and emit chemical substances. All of these pollutants are known to affect animals. Most studies on anthropogenic pollution address the impact of pollutants in unimodal sensory domains. High levels of anthropogenic noise, for example, have been shown to interfere with acoustic signals and cues. However, animals rely on multiple senses, and pollutants often co-occur. Thus, a full ecological assessment of the impact of anthropogenic activities requires a multimodal approach. We describe how sensory pollutants can co-occur and how covariance among pollutants may differ from natural situations. We review how animals combine information that arrives at their sensory systems through different modalities and outline how sensory conditions can interfere with multimodal perception. Finally, we describe how sensory pollutants can affect the perception, behaviour and endocrinology of animals within and across sensory modalities. We conclude that sensory pollution can affect animals in complex ways due to interactions among sensory stimuli, neural processing and behavioural and endocrinal feedback. We call for more empirical data on covariance among sensory conditions, for instance, data on correlated levels in noise and light pollution. Furthermore, we encourage researchers to test animal responses to a full-factorial set of sensory pollutants in the presence or the absence of ecologically important signals and cues. We realize that such approach is often time and energy consuming, but we think this is the only way to fully understand the multimodal impact of sensory pollution on animal performance and perception.     

EFFICIENCY OF A CYCLIC DESIGN AND A MULTIDIMENSIONAL SCALING SENSORY ANALYSIS TECHNIQUE IN THE STUDY OF SALT TASTE

Multidimensional scaling (MDS) procedures produce maps of stimulus points, the dimensions of which may be sensory and/or physicochemical attributes. Our objective was to test the efficiencies of a cyclic design and a MDS method in the study of saltiness perception. Two levels of NaCl (0.1% and 0.2%, w/v) were added to two nonionic and two ionic gum solutions prepared to two concentrations. Subjects scaled dissimilarities between pairs of solutions and rated each sample for flavor attributes. Solution pairs were selected using cyclic designs. Repeated measures ANOVA determined that added NaCl was the only significant factor affecting saltiness perception. In contrast, from KYST-2A MDS maps, we concluded that saltiness perception was influenced by gum property, gum concentration, presence of calcium and potassium, and related to binding of the sodium ion (Na⁺) as determined by ²³Na nuclear magnetic resonance (NMR) spectroscopy. The MDS cyclic design successfully reduced the number of samples and subjects when using experienced subjects.     

Task-induced neural covariability as a signature of approximate Bayesian learning and inference

Perception is often characterized computationally as an inference process in which uncertain or ambiguous sensory inputs are combined with prior expectations. Although behavioral studies have shown that observers can change their prior expectations in the context of a task, robust neural signatures of task-specific priors have been elusive. Here, we analytically derive such signatures under the general assumption that the responses of sensory neurons encode posterior beliefs that combine sensory inputs with task-specific expectations. Specifically, we derive predictions for the task-dependence of correlated neural variability and decision-related signals in sensory neurons. The qualitative aspects of our results are parameter-free and specific to the statistics of each task. The predictions for correlated variability also differ from predictions of classic feedforward models of sensory processing and are therefore a strong test of theories of hierarchical Bayesian inference in the brain. Importantly, we find that Bayesian learning predicts an increase in so-called “differential correlations” as the observer’s internal model learns the stimulus distribution, and the observer’s behavioral performance improves. This stands in contrast to classic feedforward encoding/decoding models of sensory processing, since such correlations are fundamentally information-limiting. We find support for our predictions in data from existing neurophysiological studies across a variety of tasks and brain areas. Finally, we show in simulation how measurements of sensory neural responses can reveal information about a subject’s internal beliefs about the task. Taken together, our results reinterpret task-dependent sources of neural covariability as signatures of Bayesian inference and provide new insights into their cause and their function.     

IMPORTANCE OF SENSORY FACTORS IN PROCESSED SEAFOOD: METHODS AND RESULTS1

This paper considers “importance” of sensory attributes from the consumer point of view, with emphasis on processed seafood. There are three key measures of importance. Attitudinal importance measures what consumers think to be important and refers to general opinions about the category. Sensory system importance measures how strongly different sensory inputs (e.g., appearance, aroma, taste, texture) “drive” overall liking, and show the key sensory inputs to which consumers attend. Attribute level importance measures the relation between sensory intensity and overall liking for each attribute. It shows which specific attributes drive liking, how liking varies with the specific sensory attribute, and whether (and at what sensory level) there exists an optimal level.