The Objective Situation As a Determinant of Consumer Behavior

Russell W. Belk, University of Illinois
[ to cite ]:
Russell W. Belk (1975) ,"The Objective Situation As a Determinant of Consumer Behavior", in NA - Advances in Consumer Research Volume 02, eds. Mary Jane Schlinger, Ann Abor, MI : Association for Consumer Research, Pages: 427-438.

Advances in Consumer Research Volume 2, 1975      Pages 427-438


Russell W. Belk, University of Illinois

[Portions of this research utilize data supplied by Leo Burnett, U.S.A.. The cooperation of Stuart Agres in making these data available is gratefully acknowledged.]

[Russell Belk is an assistant professor in the College of Commerce and Business Administration, University of Illinois, Urbana, Illinois 61801.]

Because the influence of situations is so pervasive, the notion that consumer behavior depends upon the situation is as easy to embrace as it is difficult to extricate. The development of taxonomies of consumer situations is discussed as one essential step in obtaining an understanding of the role of situations in determining consumer behavior. A means for developing such taxonomies based on similarities in behaviors elicited by objective situations is proposed and illustrated. Additional issues concerning the relationship between situations and behavior are outlined, and the relevance of objective characterizations of situations to the study of these issues is discussed.

The potential of situations to affect behavior presents a challenging premise for a new approach to the study of consumer behavior. There is now ample evidence and speculation that consumer behavior is a function of the interaction between the individual and the situation. [Speculation on this point is clearest by Ward and Robertson (1973), Engel, Kollat, and Blackwell (1969), and Tucker (1964). Several seminal studies have found the influence of situational interactions to be more sizeable than the variance attributable to the influence of individual differences alone (Sandell, 1968; Bishop and Witt, 1970; Belk, 1974).] There is convincing reason to suspect that substantially more behavioral variance may be accounted for by introducing situational variables into research than by ignoring or controlling them. [This hypothesis is supported by the studies cited in footnote 3 and in studies by Hansen (1972), Sheth (1971), and Endler and Hunt (1969).] And there is adequate justification for the conclusion that current explanations and predictions of consumer behavior offer much room for the improvement which may be provided by specifying situational conditions. What there is not, is any agreement as to just how the delineation and inclusion of situations in consumer research should proceed. It is the thesis of this paper that the objective characterization of consumer situations is a vital link in any such research process. After developing and illustrating this perspective, its place in an overall scheme for investigating situational influence on consumer behavior is elaborated.


A situation is most simply defined by a locus in time and space. At this most basic conceptual level there is widespread, although not universal, consensus. [A primary exception is Barker's (1968) "behavioral setting" which adds a behavioral dimension to this locus. By using a behavioral sequence or "action pattern" to help in defining boundaries for the situation, Barker extends the time and place dimensions to potentially longer and larger units than those considered in the following discussion. By extending all three dimensions to still larger units, we approach the concept of environment, as it is coming to be used (see Proshansky, Ittelson, and Rivlin, 1970).] As Hansen (1972) points out, the situations of greatest concern in consumer behavior occur at times and places of communication, purchase, and consumption. Agreement also seems likely on this position. But beyond drawing these very general boundaries and categories for situations of interest, the conceptualization of situation is vague. W. I. Thomas began researching situation 50 years ago under a definition of situation as a configuration of factors conditioning a behavior reaction (Thomas, 1927). His work has since been criticized for giving rise to studies examining such wide ranging factors of "situation" as social institutions, groups, neighborhoods, individual experiences, events, and even complete illusions (Volkart, 1951). Other researchers have restricted their definitions of situation to include only role expectations and expected reactions of others (Warner and DeFleur, 1969), only geographical and architectural surroundings (Golledge and Zannaras, 1973), or only educational background (Nasiter, 1968). In attempting to operationalize Rokeach's concept of situation, Roman (1969) has even used "situation" to mean brand of product. This represents an extreme flexibility in interpreting a term which Rokeach (1967) has used to mean stimuli as broadly conceived as events, activities. and social systems.

Rather than seek to resolve such differences with a highly specific and artificial distinction as to what constitutes a situation, it is more useful at this point to view situation as "all those factors particular to a time and place of observation which do not.follow from a knowledge of personal (intraindividual) and stimulus (choice alternative) attributes, and which have a demonstrable and systematic effect on current behavior" (Belk, 1974). This definition specifies that the person, the situation, and the stimulus object are distinct sources of influence on behavior, and parallels such classifications as Helson's separation of organic, contextual, and focal cues in perception (Helson, 1964). Moreover the definition stipulates that of the nearly infinite number of cues surrounding a person at a given time and place, the situation of concern includes only those features which can be shown to influence behavior. Under this stipulation only those cues within the individual's immediate sphere of sensation need be considered as potential components of the situation. Within the framework of these minimum criteria for a situation, research may empirically answer the question of what elements constitute a consumer situation, according to the ability of these elements to alter consumer behavior.

Specifying Potential Elements of a Situation

In order to determine the situational conditions which do influence behavior, it is first necessary to specify the situational conditions which might influence behavior. To illustrate, suppose we randomly selected a person who is at this moment entering a department store in one of the major urban areas of the world. The task of specifying the situation is clarified to some degree by first deciding which potential determinants of behavior should be attributed to the person and stimulus objects. The person may be characterized by observable features (e.g., sex, approximate age, weight, and height) as well as unobservable features (e.g., personality, intellect, occupation, and skills). Such features may be regarded as situation-free person characteristics. For example, if the person were observed in a restaurant instead of a department store, such characteristics would not be affected. Of the many potential stimulus objects in the department store, suppose this person is now standing in front of a counter containing gloves. Because the gloves are now potential objects for behavior, they may be removed from the situation and treated as separate source of influence. The characteristics of the gloves which may affect behavior are only those which this person is able to perceive directly (e.g., color, size, labeling, materials, and textures). Any added significance for a pair of gloves derives from an interaction of these characteristics with those of the individual or the situation.

All remaining potential behavioral influences in the above scenario are a part of the potential situation. We may externally note the most obvious elements of the situation in such features as temperature, time of day, persons present, transactions taking place, sounds, odors, and decor. Like stimulus object attributes, these physical qualities are variously perceived by the individual, but the situation itself may be described independently of these interPretatiOnS. There may however be other stimuli specific to this time and place which reside within the individual and have no external correlates present. These unobservable features involve such momentary or episodic states of the individual as moods, plans, and purposes. For instance, it may well be important to know whether the department store shopper is taking a few moments while on a lunch break to select a gift for the birthday of a family member, or whether the person is in the store to get out of the rain, return a defective product, or shop for another general or specific type of product. Inasmuch as such momentary internal states may potentially affect behavior, they fall within the scope of the minimum criteria for a situation as long as they may not be construed as the result of an interaction between the characteristics of the individual and those of the stimulus object or the physical features of the situation. Thus a mood created by the background music in the store would be a response rather than a situational stimulus, while a mood created by events outside of the present physical situation may be regarded as a part of the current situational stimuli.

The Objective Situation

Recognition of both external and internal aspects of the situation poses a major stumbling block to situational research. Mausner (1963) states the problem must succinctly: "If one specifies the stimulus in terms of the nature of the receiver, lawfulness becomes impossible. But without knowledge of the state of receiver one cannot even begin to state the nature of the antecedent conditions which may be expected to lead to behavior" (Mausner, 1963, p. 107). For instance, the explanation that a person refrained from buying a new brand because the situation was a threatening one, is tautological if both the behavior and the situation are viewed as entirely idiosyncratic to the person. It tells us no more than the comparably hollow individual differences explanation that this is the sort of behavior engaged in by the sort of person who engages in this sort of behavior. Both explanations concentrate on the person and the response, and neglect the antecedent conditions for the behavior. While the inclusion of completely observable situational characteristics offers a chance for improved explanability, in this case such a description would fail to disclose the factors which make the situation threatening unless this condition would always result from the presence of the same set of physical stimuli. If instead the situation is threatening because of an internal mood or purpose of the individual, the richness added to the specification of situation by a knowledge of this state is lost.

There is a way in which we can maintain much of the potential richness of the concept of situation while still maintaining an essentially external perspective. To do so requires that the situation be operationally defined in "objective" terms. An objective element of the situation is one which is capable of external verification without the need to construct measures of internal states of the individual. In addition to the physical features of the situation, objective descriptions may include the existence of external facts and events which bear upon current behavior even though they are not themselves physically a part of that situation. For example, if it can be determined that the stimulus object is an item about which friends or family have recently raised complaints, or is an item to be served at an upcoming dinner party for the new boss, these objective descriptions can be included in the situational specification directly rather than attempt to measure whether or not the situation is regarded to be threatening by the individual participant. Similarly we might ascertain that the individual has had a hectic day with the children, or has just finished the last of several difficult final examinations, or has just gotten a promotion, instead of attempting to measure the mood this person brings to the situation. Besides the potential for increased explanatory richness compared to completely physically conceived situations, objective situational specifications allow greater accuracy in description and control of the situation than is possible using internal specifications.


The discussion to this point has been concerned with a more precise specification of what is meant by the term "situation." It has been argued that a rich and meaningful use of this concept includes both the antecedent conditions for the momentary individual states which a person brings to a given time and place, and the physical features which he finds there. If we were able to fully enumerate the objective situations which this concept entails, we would doubtless conclude that no two situations are absolutely identical. Since the same may be said of consumers, the task of understanding the role of situations in determining behavior is no more or less imposing than that of understanding the role of consumer characteristics in determining behavior. The study of both of these sources of influence on consumer behavior may be aided substantially by the development of taxonomies which systematically classify their respective domains. However inadequate they may be, taxonomies of individual differences do exist in the form of psychological and behavioral classifications of individuals or attributes of individuals. There have been few attempts to develop comparable classifications of situations, [The several notable attempts to develop comprehensive taxonomies of situations are those by Sherif and Sherif (1956), Sells (1963,1969), Bellows (1963), Allen (1965), and Kasmar (1970).] and no existing taxonomy is especially appropriate for describing consumer behavior situations. There is however a method of developing limited taxonomies of situations which is especially germane to the classification of consumer behavior situations. This method recognizes that rather than classifying situations on the basis of shared attributes, it is more directly relevant to classify situations on the basis of their ability to elicit similar behaviors. Such an approach simultaneously helps to reduce and group the many situations in which consumer behavior takes place, and does so in a way which fulfills the definitional criterion that a situation include only those elements which can be shown to systematically affect behavior.

A Method for Behavioral Taxonomy of Situations

The starting point for classifying situations via their tendency to elicit similar behaviors is a three dimensional data matrix where the dimensions are persons, situations, and behaviors. The data within this matrix are observations of whether or not a given behavior occurs for a given person in a given situation. Although this observational approach for collecting data for the matrix has been demonstrated to be feasible (Fredericksen, Jensen, and Beaton, 1972), a projective approach which asks subjects how likely they would be to engage in various behaviors in each of various situations has been more common. In either case, the three dimensional matrix which results may be reduced to a more general form in several ways. The traditional solution would be to collapse or average over one dimension to provide a two-way table on which to derive correlations for input to factor or cluster analysis. In order to classify situations in this manner the matrix would be averaged over subjects, and situations would be correlated across behaviors. In order to classify behaviors the matrix would be averaged over situations, and behaviors would be correlated across subjects. Factor analyzing these sets of correlations would yield types of situations and types of behaviors, but it would not yield insights into the relationship between the two sets of factors obtained. That is, no description of types of behavior evoked by types of situations would result. In order to deal with this problem, Levin (1963) first demonstrated the application of Tucker's three-mode factor analysis to behavioral taxonomy (Tucker, 1963). Three-mode factor analysis of the sort of data suggested here not only provides classifications of types of behavior evoked by types of situations, it also distinguishes types of individuals for whom elicited behaviors show different patterns over the set of situations. The procedure simultaneously extracts factors in each of the three modes of the matrix, and describes their interrelationships in terms of a core matrix. By multiplying the entries in this three dimensional core matrix by the entries in the three factor loading matrices before rotation or factor deletion, the original data matrix may be reconstructed. Summary details of this technique may be found in Vavra (1972a,b) and Belk (1973,1974).

Taxonomy of Objective Consumer Situations -- An Example

In order to apply three-mode factor analysis to the development of situational taxonomies, an illustrative set of consumer situations and choice behaviors were prepared concerning patronage of fast food and take-out restaurants. [Data were provided by Stuart J. Agres, Manager of Research Development, Leo Burnett.] Ten situations and ten alternative choices were presented to subjects who responded on a six point scale from "Not at All Likely" to "Very Likely" to indicate the probability that each choice would be made in each situation. Subjects were 98 married females in a single community. Their responses to each stimulus and response pair were standardized, and the transformed data were analyzed using three-mode factor analysis.

The situations employed and their loadings on the four factors derived in this mode are shown in Table 1. The selection of situations for investigation cannot hope to be representative until further applications of behavioral taxonomies have been accumulated. The situations used in this application may be seen to be specified in objective terms, with more emphasis placed on the states the individual brings to the situation than on the physical features he finds there. Based on the factor loadings in Table 1, it appears that the first situation factor may be described as variety-seeking situations. The second -factor involves situations in which the subject is entertaining guests. Factor three is clearly picnic situations. And the final factor seems to represent occasions which are relaxation-seeking and informal. The last factor tends to include situations which occur at the end of a long day or upon completion of some involving activity. Interestingly, whether these antecedent conditions tend to be positive (a pleasant chat; having too good a time) or negative (too tired) their effect on choice behavior is similar.

The other modes analyzed provided four factors for responses and four factors for persons. The response factors described 1) multi-course dinners, 2) fast food A, 3) fast food B, and 4) self-prepared hot dogs or sandwiches. The relationship of these choices to the four types of situations is shown in Table 2. This table is the portion of the core matrix corresponding to one of the four types of persons, and the entries reflect the relative likelihood of choosing each alternative meal type under each alternative situation type for this subgroup of people. The more negative entries indicate unlikely choices and the more positive entries reflect likely choices. The type of person depicted is most likely to select takeout food A in variety-seeking situations and relaxation-seeking situations, and takeout food B for picnics and when entertaining guests. Others of the four person types derived showed markedly different patterns of behaviors. The multi-course dinners tended to be unlikely choices for all groups, which is consistent with the observation that these were generally less popular fast food outlets in the community. Twenty of the 98 subjects were classified as belonging to the classification of persons shown, and they were further characterized by their higher age, fewer children, and lower income than those in the other three groups.



This example serves to suggest the potential uses of behavioral taxonomies for situations involving a particular product category. Obviously it would not be suggested that the taxonomy derived here will necessarily describe relevant situations in other product or behavior categories. Of the four types of situations found to affect choices in this instance, only the relaxation-seeking factor finds parallels in similar studies for choice of snack products, meat products, and motion pictures (Belk, 1973,1974). The technique of behavioral taxonomy, however, is highly generalizable. It presents a useful and versatile means of screening potential stimulus configurations for their behavioral impact, while at the same time beginning the process of categorizing and compressing these situations. The full results of analyses like the one illustrated provide nearly complete formulas (types of choices by types of consumers in types of situations) for situational segmentation and positioning. Where the researcher has been careful to specify input situations in objective terms, the further necessary information on frequency of situational occurrences for each type of individual may be readily obtained as well.




Although obtaining behavioral taxonomies of objectively specified consumer situations is a concrete means of operationalizing the concept of situation, such studies are only the starting point for investigating situational influence. Once it has been demonstrated that a set of situations have the capability to influence consumer behavior in a certain manner, the next obvious question is why this effect occurs. In light of the core matrix from three-mode factor analysis, we may also ask why this effect occurs for some persons but not for others. To answer both questions, and especially the latter, it will be necessary to consider the internal responses which an individual makes to a situation. Examinations of the links between objective situations and internal responses are needed to reveal the ways in which the individual's thoughts, feelings, and expectations are altered by particular situations and situational cues. These internal responses, to the extent they can be measured, may then be scrutinized to form causal hypotheses about the nature of situational effects. This type of research may be expected to suggest explanations for effects which are revealed by direct situation-behavior research but nevertheless appear mysterious. For instance, the fact that both pleasant and unpleasant antecedent activity lead to similar behaviors in the prior example, may find some rationale in terms of similar expectations generated by such activities. In carrying out such research, the link between the situation as it objectively exists and the situation as it is subjectively interpreted forms an anchor to the real world which is essential if we are to avoid the sort of tautologies mentioned earlier. It is most likely that it is the absence of this crucial link in Lewin's (1935) separation of life space from physical space which has impeded the literal adoption of this perspective, despite its substantial conceptual contributions. The warning then is to begin with the situation as it exists rather than its internal representation.

Another area of concern for consumer situational research is how an individual gets to be a particular situation in the first place. Again a knowledge of internal responses to situations may help. The degree to which a given situation is attractive, avoidable, and foreseeable, should be related to its frequency of occurrence, which should in turn determine the aggregate frequency of the behaviors evoked by this situation. Research on both perception of situations (Mehrabian and Russell, 1974) and the frequency of situational occurrences (Szalai, 1966) is now under way, although the combination of these two research streams in terms of the individual's control in selecting the situations he encounters has not yet taken place. In order for a merger of these two perspectives to be possible, they must have a common denominator. Only if the situations toward which perceptions are measured are the same situations for which frequency distributions are constructed, can the two be related. In order to tie both research streams to the same objective world, the development of taxonomies of objective situations is a high priority task.

One further and more obvious contribution of objective characterizations of consumer situ&,ions involves the implications flowing from observations of situational influence on behavior. From the points of view of marketing and public policy, the relevant question is "how externally controllable are those elements of consumer situations which can be shown to influence behavior." It is apparent that the question of control is more easily assessed where the situation is measured in terms of objective elements than where the focus is on subjective interpretations. For instance a retail shopping situation described by such factors as merchandise clutter and floor surfacing has more readily apparent implications than a description of the situation by its novelty and arousal potential. Furthermore internal responses such as mood, that are only partially controllable, can be more adequately evaluated for controllable (e.g., color) and uncontrollable (e.g., weather) factors when they are specified in objective terms.


The clearest means of defining consumer situations resides in the objective features which characterize a locus in time and space. Such elements include both antecedent conditions for the internal states the participant brings to the situation and the physical elements he attends there. These features lend themselves to direct behavior taxonomy without the need to infer internal responses of the individual to the situation. While objectively defined situations are only one component of the necessary perspectives for researching situations, the objective situation is the common thread which is required of any study of situations if it is to add to our ability to explain, predict, or alter consumer behavior.


Allen, B. L. Situational factors in conformity. In L. Berkowitz (Ed.), Advances in experimental social psychology. Vol. 2. New York: Academic Press, 1965.

Barker, R. G. Ecological psychology: Concepts and methods for studying the environment of human behavior. Stanford: Stanford University Press, 1968.

Belk, R. W. Application and analysis of the behavioral differential inventory for assessing situational effects in consumer behavior. In S. Ward and P. Wright (Eds.), Proceedings. Urbana, Ill.: Association for Consumer Research, 1973, pp. 370-380.

Belk, R. W. An exploratory assessment of situational effects in buyer behavior. Journal of Marketing Research, 1974, 11, 156-163.

Bellows, R. Toward a taxonomy of social situations. In S. B. Sells (Ed.), Stimulus determinants of behavior. New York: Ronald Press, 1963.

Bishop, D. W. and Witt, P. A. Sources of behavioral variance during leisure time. Journal of Personality and Social Psychology, 1970, 16, 352-360.

Endler, N. S. and Hunt, J. McV. S-R inventories of hostility and comparisons of the proportions of variance from persons, responses and situations for hostility and anxiousness. Journal of Personality and Social Psychology, 1969, 9, 309-315.

Engel, J. F., Kollat, D. T., and Blackwell, R. D. Personality measures and market segmentation. Business Horizons, 1969, 12, 61-70.

Fredericksen, N., Jensen, O., and Beaton, A. Prediction of organizational behavior. Elmsford, N.Y.: Pergamon Press, 1972.

Golledge, R. G. and Zannaras, G. Cognitive approaches to analysis of human spatial behavior. In W. H. Ittelson (Ed.), Environment and cognition. New York: Seminar Press, 1973.

Hansen, E. Consumer choice behavior. New York: The Free Press, 1972.

Helson, H. Adaptation-level theory. New York: Harper and Row, 1964.

Kasmar, J.V. The development of a usable lexicon of environmental descriptors. Environment and Behavior, 1970, 2, 153-169.

Levin, J. Three-mode factor analysis. (Doctoral dissertation, University of Illinois) Ann Arbor, Mich.: University Microfilms, 1963. No. 6099.

Lewin, K. A dynamic theory of personality. New York: McGraw-Hill, 1936.

Mausner, B. M. The specification of the stimulus situation in a social interaction. In S. B. Sells (Ed.), Stimulus determinants of behavior. New York: Ronald Press, 1963.

Mehrabian, A. and Russell, J.A. An approach to environmental psychology. Cambridge: The MIT Press, 1974.

Nasiter, D. A note on the contextual effects and the political orientation of university students. American Sociological Review, 23, 1968, 210-214.

Proshansky, H. M., Ittelson, W. H., and Rivlin, L. G. Environmental psychology. New York, Holt, Rinehart, and Winston, 1970.

Rokeach, M. Attitude change and behavior change. Public Opinion Quarterly, 30, 1967, 529-550.

Roman, H. S. Semantic generalization in formation of consumer attitudes. Journal of Marketing Research, 1969, 6, 369-373.

Sandell, R. G. Effects of attitudinal and situational factors on reported choice behavior. Journal of Marketing Research, 1968, 4, 405-408.

Sells, S. B. Dimensions of stimulus situations which account for behavioral variance. In S. B. Sells (Ed.), Stimulus determinants of behavior. New York: Ronald Press, 1963.

Sells, S. B. Toward a taxonomy of organizations. In W. W. Cooper, A. J. Leavitt, and M. W. Shelly II (Eds.), New perspectives in organization research. New York: Wiley, 1969.

Sheth, J. N. Canonical analysis of attitude - behavior relationship. Paper presented at TIMS eighteenth international meeting, 1971.

Sherif, M. and Sherif, C. W. An outline of social psychology. Rev. Ed. New York: Harper and Row, 1956.

Szalai, A. The multinational comparative time budget research project. American Behavioral Scientist, 1966, 10, 1-12+.

Thomas, W. I. The behavioral pattern and the situation. Proceedings, Twenty-second annual meeting, American Sociological Society, 22, 1927, 1-13.

Tucker, L. R. Implications of factor analysis of three-way matrices for the measurement of change. In C. W. Harris (Ed.), Problems in measuring change. Madison: -University of Wisconsin Press, 1963.

Tucker, W. T. The social context of economic behavior. New York: Holt, Rinehart, and Winston, 1964.

Vavra, T. G. An application of three-mode factor analysis to product perception. Proceedings, American Marketing Association fall conference, 1972a, Chicago: American Marketing Association, 578-583.

Vavra, T. G. Factor analysis of perceptual change. Journal of Marketing Research, 1972b, 9, 193-199.

Volkart, E. H. Introduction: Social behavior and the defined situation. In E. H. Volkart (Ed.), Social behavior and personality. New York: Social Sciences Research Council, 1951.

Ward, S. and Robertson, T. S. Consumer behavior research: Promise and prospects. In S. Ward and T. S. Robertson (Eds.), Consumer behavior: Theoretical sources. Englewood Cliffs, N.J. Prentice-Hall, 1973.