Empirical Unit of Acceptance Behavior

Food acceptance is a single momentary overt act of a person (or other organism) on a particular occasion. Thus, any acceptance of a food occurs in a particular context of alternative foods, bodily condition, physical and social environment, and other aspects of mental state. Also, any measure of behavior toward food that results from more than one act by one person toward one food in one situation might be confounded by variations over time in that person or by differences among people, foods, or situations.

A. Choice and Relative Acceptance. When two or more foods are present, the acceptance of one of them is an overt choice among the foods. In this context on this occasion, the chosen food has been preferred over the other food(s). Repetition of the choice by that person in similar contexts can establish that the choice was not random, or in some circumstances this can perhaps be assumed on other grounds.

Even if only one food is ever present at one time (i.e., monadic presentation), opportunities to accept can be provided in similar contexts for two or more foods in succession. Then, from the acceptance or refusal of the different foods, we can construct an observed relative acceptance.

This combination of acceptance and refusal on separate occasions can be represented by the same parameter as the combination of acceptances and refusals among the same set of foods presented simultaneously on repeated occasions (e.g., the proportion of acceptances of one food to the total acceptances from the set). Qualitative similarity of successive and simultaneous relative acceptances is evidence for the same underlying preference structure. Quantitative similarity is unlikely, however, since the strongest preference is freer to dominate behavior in the simultaneous test. The successive (i.e., monadic) test therefore has the advantage of sensitivity to less than the strongest disposition to accept a food. However, the real purchase or eating situation might be one of choice.

The issue that arises is always crucial: What are the objectives of the investigation? In particular, what situation of acceptance are we seeking to understand? In practice, using a familiar range of food products or variants of a food and monadic presentations within a fairly naturalistic situation provides sufficiently realistic and precise data to characterize and even to quantitate influences on either multi-item choice or singleitem acceptance/rejection in the normal context.

B. Degrees of Acceptance and Graded Influences. The discussion so far has been couched in terms of acts of acceptance or rejection of particular foods in a particular situation. An acceptance is categorical—it occurs or it does not (except perhaps when two or more foods are put into the mouth or the shopping basket at once). Degrees of acceptance would be quantitatable as frequencies of acceptances (or refusals) in repeated tests.

Nevertheless, a single act can be used to generate an estimate of such frequencies over many yes/no tests. With people the simplest way to do this is to ask for a verbal judgment of the frequency with which they would accept a monadically presented item in some defined or assumed context. Alternatively, they can be asked to express a degree of liking or pleasantness; this appears to be more introspective, but that is exposed as an illusion of language when the hedonic rating is anchored on acts in specified situations. Then these ratings become the same as acceptance frequency ratings. In any case the qualitative validity and quantitative calibration of acceptance ratings of any sort are based on the actual frequencies of acceptance in the real-life decisions.

An influence on acceptance might also be inherently categorical (i.e., either present or absent), having no intermediate strengths. Brand name and food type are examples. However, these categories could be resolvable into sets of perceptible characteristics of the brand image or the food type’s composition or uses; in other words, brand names or food types might scale onto several continua, whether sensory (e.g., sweet), somatic (e.g., quick filling), or social (e.g., breakfasttime or preschool age). The basis can be identified for either a categorical difference or a metric continuum by finding the observable characteristic(s) of an influence on acceptance that regresses linearly and precisely onto degrees of acceptance by the individual. This regression provides a measure of sensitivity to that influence, namely, the just discriminable difference in the influence by that response.

C. Generalizing. The usual concept of the acceptability of a food is highly abstract and often rather unrealistic. Food acceptability, as commonly conceived, generalizes across consumers, eating or purchasing contexts, and sets of foods among which choices are being made. In consequence, sensory acceptability or palatability is commonly misconstrued as an inherent characteristic of the food. Even if the sensory influences on acceptance of a food did not vary among people or situations (as they commonly do), palatability would still be a characteristic of behavior toward the food, not of the food itself—of a causal relationship between a person and a food, not of one of the terms in that relationship.

At a moment in a meal, whether or not more of a food will be eaten depends on how boring and filling it has so far become and on anticipatory decisions when serving onto the platter or when purchasing the items for the meal. Whether another food now becomes accepted could depend on any of these factors for other available foods and on the next food’s suitability to the upcoming stage of the meal. In Western cuisines the acceptabilities of meat pie and apple pie reverse between the first course and the dessert, although, of course, they do not compete when shopping for an entire meal.

Acceptance of a food is often also contingent on the time of day (cf. breakfast foods, between-meal hot drinks, and snack items), company at the table (e.g., spouse, children, and guests), season of the year (e.g., hot or cold weather), health concerns, specialness of the occasion, and so on. Still further factors can be particularly influential at the point of purchase (e.g., price, advertising messages, and package design and information). Thus, a generalization about food acceptance must either specify the particular context assessed, in all its potentially influential aspects, or include the effect of an influence within the generalization.

Most challenging of all, people differ from one another in how they put all of these sensory, somatic, and social influences together into the habitual acceptance decisions. Until recently, all approaches to influences on food acceptance have assumed that there are simply additive combinations among influences, and that therefore the analysis of data lumped across people would not be qualitatively, or even quantitatively, too misleading. In fact, many influences simply do not add to each other. Instead, the distances of distinct influences from their best accepted level are combined in multi-dimensional space according to the Pythagorean Theorem (square root of sum of squares).

Furthermore, qualitatively different structures of interactions among influences can be quite common. Indeed, there has long been evidence that the results of sensory tests with and without brand names sometimes differ greatly; that acceptabilities of food products after use often differ from, and can be even more diverse than, acceptabilities at first acquaintance and that the awareness parameter in a brand life-cycle model has nothing to do with the repurchase parameter.