Food and Nutrition, Perception. Sensory Measures

Sensory measurement as it relates to foods is complicated by the fact that there are two levels at which assessment is required. The first concerns the sensory properties of foods; the second concerns the sensory capacities of individuals. Sensory properties of foods consist of attributes such as texture, color, and odor as determined by each item’s chemical composition.

Whether these characteristics are detected and how they are interpreted are determined by the function of each individual’s sensory systems. The interaction between the two determines the perceived flavor of foods. For example, humans are polymorphic with respect to their sensitivity to phenylthiocarbamide (PTC) and related compounds. Some individuals (approximately 70% of the Caucasian American population) find low concentrations of such compounds in the food supply to have a bitter taste, whereas others fail to perceive any bitterness from the same foods. This may influence the acceptability of these items since bitterness is generally a negative taste attribute.

Study of the sensory properties of foods has generally fallen into the domain of food science. This knowledge is used for applications such as new product development, product matching, product improvement, quality control, and product ratings. A variety of tests are used for each of these purposes. Difference and sensitivity tests are designed to determine whether particular product characteristics are discernible. Judgments about whether a stimulus is present in a given context (e.g., sucrose in ketchup) or what the quality of a stimulus may be (e.g., salty for NaCl in soup) are the two most common sensitivity measures. They are termed detection and recognition thresholds, respectively, and are only statistical concepts. They are merely the level of stimulus (e.g., spice) required so that the number of correct responses (e.g., present or not present) by observers in a defined testing situation will exceed chance performance with a preset level of probability. The difficulty of the task and, as a consequence, the absolute threshold value will be determined by testing conditions.

Descriptive sensory tests are used to identify and quantify product attributes. For example, judgments may be sought for the viscosity of an array of foods with varying gelatin levels. Once again, the absolute values of responses will be determined by the testing paradigm. Viscosity ratings for a particular food following sampling of a large array of either less viscous or more viscous samples will probably be higher in the first instance than in the second.

The third class of tests, affective tests, provide insights on the palatability and acceptability of products. Sensory responses may be obtained for levels of specific product attributes (e.g., sourness of lemonade) or for the product as a whole. Although the focus is on the product, wide intra- and interindividual variability in hedonic responses necessitates the use of a well-defined panel of judges tested under highly controlled conditions. An example of an intrasubject factor is the level of hunger or satiety of the respondent. Having subjects rate a rich dessert item when hungry or full will likely lead to discrepant responses.

Differing previous experiences with foods can result in discrepant hedonic ratings between subjects. Whereas overtly bitter foods are generally regarded as distasteful in Western cultures, where exposure to such items is limited, they are considered more acceptable to groups whose customary diet includes bitter items.

Study designs for product evaluation generally involve obtaining sensory evaluations of foods or beverages with varying properties from observers with carefully controlled or at least well-defined characteristics (e.g., age, sex, educational level, economic status). This is particularly true of sensitivity and descriptive type tests. Instrumental analyses are also used, but in many instances the sensitivity of human sensory systems cannot be duplicated. Trained panelists are commonly used for these purposes. A trained panelist is an individual who, through practice and experience, is very familiar with a product or line of products and can reliably evaluate each of its attributes. It should be emphasized that these individuals are not selected on the basis of some innate supersensory capacity. Instead, they have learned to use their sensory capacities more fully. In contrast, untrained panelists are individuals with no particular expertise with a product. Often they are potential consumers of a product and are used primarily for tests of product acceptability.

Interest in the sensory capacities of humans has been particularly strong in, the field of psychology as well as the basic physical sciences (e.g., physiology, chemistry, anatomy). Measurements are again generally made in the three domains of sensitivity, description, and affect. However, because of the difference in orientation (i.e., focus on the respondent instead of the stimulus), testing procedures often vary. Test stimuli are commonly held constant and individual variation in responses to these stimuli are the measures of interest.

Individual differences in sensory function may be attributable to both innate physiologic as well impinging environmental or external factors. An example of one innate physiologic influence is an individual’s genetic constitution, which determines their sensitivity to the odors of different compounds. An undetermined number of individuals in the population have specific anosmias, the inability to perceive the odor of low concentrations of certain compounds (e.g., androstenone—the odorant that attracts swine to truffles). Since the perceived odor of a food is comprised a combination or profile of the volatile compounds that it releases, the absence of one constituent part will alter the overall quality of the food’s odor.

External influences on sensory function include factors such as state of hunger, health status, and smoking habits. As these factors are in a constant state of flux, their influence and, as a consequence, the sensory function of an individual can vary widely over time. For example, nasal congestion may reduce olfactory (but not gustatory) sensitivity, smoking diminishes taste sensitivity acutely, and being hungry can enhance the pleasantness of tastes and smells.

Preference is the sensory attribute most strongly associated with food selection and ingestion. Preference responses are obviously based on discernment of product characteristics, but measures of sensitivity and intensity hold little predictive power with respect to ingestive behavior. They appear to be translated into an integrated hedonic message that is used to evaluate the acceptability of the item. Given the complexity of information used to derive such a judgment,it is not surprising that measurement of preferences is similarly complex. Responses to questions about the preferred frequency of ingestion of a food, preferred level of an attribute in the food, or preference for foods with a given profile of attributes can, and typically do, provide different impressions.

For example, an affectively neutral food (e.g., bread, milk, butter) may be ingested frequently whereas a highly preferred item (e.g., pumpkin pie, lobster, caviar) may be ingested only on special occasions. A food with a high level of a given component (e.g., salt on potato chips) may be highly preferred, but other items with high levels of the same component can be disliked (e.g., anchovies). Only when multiple dimensions of taste preferences are assessed can predictions be made about consummatory behavior.