The Context of Social, Cultural, and Consumer Behavior


Gordon A. Haaland (1974) ,"The Context of Social, Cultural, and Consumer Behavior", in NA - Advances in Consumer Research Volume 01, eds. Scott Ward and Peter Wright, Ann Abor, MI : Association for Consumer Research, Pages: 145-153.

Advances in Consumer Research Volume 1, 1974    Pages 145-153


Gordon A. Haaland, University of New Hampshire

[Associate Professor, Department of Psychology, University of New Hampshire.]

This paper suggests a context for theory and research in social and consumer behavior. While these behaviors are not seen as identical, they are often related to similar constructs such as learning, attitudes, influence and reference groups. In addition, social and consumer behavior are conceptualized as problems to be understood at varying levels of complexity, from individuals acting alone to large groups of people behaving in some organized way.

The analysis proposed here involves delineating levels of complexity as systems of related phenomena. It is further suggested that these levels of complexity, called levels of analysis, provide a basis for understanding some of the successes and failures in research and theory on social and consumer behavior and, when related to cross cultural analysis, present (1) criteria for the development of principles of behavior which can be considered cross cultural and (2) criteria for the prediction of culture specific behavior.

Social and behavioral scientists are accustomed to thinking about their respective disciplines in one of several ways:

1. Some see distinct subfields of a theoretically integrated whole. The theory which can integrate all social phenomena has yet to be developed.

2. Others observe rather indistinct subfields of an overall science dealing with social organisms. A theoretical integration is considered possible but the current division into disciplines is considered quite arbitrary.

3. Some conceive of social science as a single discipline studying different problems. The distinction between subfield and problem is only one of degree. A topic is considered a problem if it is immediately reducible to a standard set of propositions which are drawn from some overall theoretical stance. A subfield denotes a set of concepts or ideas which are theoretically reducible to a single proposition or a set of necessarily related propositions.

4. Finally, social science can be viewed as subfields and problems, related only in that the source of data is similar, i.e., people. The proposition of this paper is that systematic consideration of this last point provides the greatest potential for meaningful development in both method and theory.

A word about the concept of paradigm is in order. Although Kuhn (1962) implies that social science needs a paradigm like that of natural science, i.e., one which would give a common theoretical base to all pursuits in social science, his point only holds if one starts by assuming that all social behavior is philosophically reducible to a single set of related propositions. The first three characterizations mentioned above are in essential agreement in supporting this point. If one assumes that social behavior is not a unitary phenomenon, however, then any of the four approaches can be paradigmatic. It would be necessary only to state the range of phenomena for which the paradigm was developed and distinguish these phenomena meaningfully from other phenomena.

Social behavior can be most meaningfully understood by conceptualizing several discrete levels of analysis. These levels need not be conceptually or methodologically related to each other nor would one level be reducible to another, i.e., explained in terms of propositions developed at another level or analysis. There are numerous implications derived from approaching the study of social behavior in this way, as shown in the propositions discussed below.

Levels of Analysis

There have been other forms of the concept of levels of analysis suggested by social theorists. Various social theorists, particularly systems theorists, have made an effort to provide a perspective on understanding social behavior.

In a recent paper on psychology and culture, Triandis, Malpass, and Davidson (1973) have suggested a set of concepts which represent a collectively exhaustive set of variables for cross cultural comparison. They suggest that behavior is a function of a person's abilities, subjective culture, personal dispositions, physical environment, social structure and other persons. Ability, subjective culture, and personal dispositions might be considered cognitive variables related to the individual's functioning as a psychological organism. These variables are contemporaneous; that is, they are always present and affecting the individual's behavior. They are, in effect, the individual. The other variables can be considered as more historical or contextual. For research purposes, the physical environment, social structure, and interaction with other people can be conceptualized and varied independently of the individual. At any given moment in the explanation of social behavior, however, one would expect all of these variables to be acting. While we can separate these variables for research purposes, they collectively represent the range of influences on behavior and are always present as factors relevant to an individual's behavior.

The delineation of these variables points up the basic dilemma addressed by this paper. How can social science develop sets of variables or concepts which are considered always to be acting on the individual and at the same time attempt to analyze some behavior as if some of the variables are not at the moment relevant? For example, the individual is never without a culture even though the researcher's basic interest is in his learning.

A few examples of the problems raised by levels of analysis might be useful. These examples are cross cultural in that they are drawn from the author's experience of living in Norway for a year. They are comparisons only in the informal sense of comparison with experiences here. The point of these examples is that an understanding or analysis of the behavior involved can occur at several levels of complexity.

There is a town in the western part of Norway called Mongstad. This is of current interest to social scientists because a large industrial concern has moved in during the past few years and has caused considerable disruption. Most of the residents have moved either because their land and houses were in the path of development or because all their neighbors had moved. Research is being conducted on the attitudes of the residents, the changes in patterns of interpersonal behavior, changes in the governmental style of the town, the personal and structural adjustments made by the residents, and even some concern over how a larger segment of the society views the plight of these people. The question is, how is the data which is relevant to personal behavior to be related to the data on the town as a system, if at all? Can such information be handled within the rubric of a singular theory or would an analysis of both personal and societal levels be appropriate and meaningful even if these levels were not conceptually related?

Another example of interest is the experience of being treated as a cultural stereotype. Shortly after my arrival in Norway I noted that some friends and colleagues would behave in a way which indicated they knew exactly how I would behave in a particular situation. The problem was that I had no intention of behaving that way and I had never seen any of them behave that way. When questioned about these expectations, they simply said "that's how an American behaves," although I had almost never seen an American behave the way they expected an American to behave. Can we use the concept of cultural stereotype to understand interpersonal behavior, international behavior, or both?

The following propositions are meant to address these questions:

1. The study of social behavior is a systematic examination of any of several levels of analysis (society, organization, group as system, individual in group).

2. These levels have meaningful boundaries.

3. Each level of analysis has its own context or behavior setting (time and place).

4. The selection of a level dictates methodological factors and design questions.

5 Each level may generate its own theoretical statement. Different processes may be used to understand behavior of different levels.

6. This theoretical statement may be unrelated to and even have paradoxical assumptions to explanations at other levels.

7. The observation that the same behavior may be examined at different levels and that behavior always involves people is simply a confusing accident.

8. Other levels, in which the level under analysis is obviously embedded, have no necessary relevance for conceptualizing behavior at the level under analysis.

9. The selection of a level of analysis as the most appropriate or fruitful for social analysis is best determined by theoretical and empirical relationships rather than by tradition or the prescriptions of other theoretical statements.

10. "Understanding" can occur by specifying relationships within a level.

11. The overall picture and theory of social behavior may require numerous levels, analyzed and portrayed independently, with possible contradictions.

Proposition one identifies the study of social behavior as a systematic examination of any of several levels of analysis. These levels include society, organizations, the groups as a system, and individuals in the group and are not collectively exhaustive levels but rather are thought to represent current work on social behavior. For example, research representing the societal level includes the typical demographic studies that are of concern in subfields of sociology or efforts to understand concepts like mobility and power in a large social system. The current work on national indicators is generally at this level of analysis. Research at this level normally describes the structure and functioning of large social systems.

Organizations as a level refers to a coherent social system which is part of a larger social system but can be identified and separated from it through the interlocking normative behaviors of the members who are part of that organization. There are rules or norms which can be abstracted from the behavior of members which provide some description of the interactive behavior of these people. Concepts such as roles, positions, and status have typically been used as abstractions for organizational description. The behavior of the members of an organization is related with respect to goals or structure of the organization although the members are not necessarily in direct communication with each other.

The group as a system refers to a small group of people whose behavior is interlocked and who are in direct communication with each other. The group as a system is conceptualized as something more than simply understanding individual behavior and generalizing to the group. As suggested by Hansen (1972) one needs concepts which are different than those dealing with individual choice behavior to deal with aggregated behavior in a consumer or social situation.

The individual in the group refers to the simplest social level, an individual in interaction with someone else. This level represents much of the traditions of social psychology and consumer behavior and includes such problems as decision making, attitude change, and social influence.

It should be noted that traditionally these levels could be organized in an hierarchical fashion. Groups as systems contain individuals who are behaving within groups. Organizations contain groups as systems and societies contain organizations. A basic premise here, however, is that this observation should have little impact on research or theory. This is a decidedly non-reductionist context for the study of social behavior.

The second proposition suggests that these levels of analysis have meaningful boundaries. This means that a level is a functional system where the phenomena (variables) that are included within the level are naturally connected to that level and have less connection to other levels. An example is cohesiveness in the level of a group as a system. This concept has meaning only in reference to groups and not to individuals, per se. Boundaries mean that the units measured are naturally related (variables) and not that the place or structure being measured dictates this relationship. It is possible for the same physical structure to yield data relevant to different levels of analysis (e.g., behavior). Berrien (1968) suggests that such boundaries are identified by some differentiation in the relationships which exist between the components inside the boundary and those relationships which transcend the boundary. Thus, behavior per se can transcend the boundary of all of these levels and ultimately is the data of social sciences. Nevertheless, the behavior that is related particularly to a level should have some special relationship to that level by speaking to the constructs of that level. There is some empirical basis for this in the research on social behavior and consumer behavior. Organization, groups as systems, individuals in groups and large social systems have appeared to be the focus of much recent research, suggesting natural levels with natural boundaries.

The third point is that each level of analysis has its own context or behavior setting (time and place). In conjunction with levels of analysis, this indicates a tentative two factor organization of social phenomena. One factor involves the different levels of analysis and the other factor is time and the environment, considered together.

Culture represents a slice across all levels of analysis at a particular time and place. This suggests that should either time or place change, culture changes. Thus, culture can help define generational differences as well as geographic differences. Culture can be conceptualized as the consideration of multiple levels of analysis at a particular time and place. The observations, however, are limited to that time and place. This idea implies that to find principles which generalize across cultures one must remain within a particular level of analysis. It is not possible to generate social theory when the variables or constructs of concern invoke more than one level. In this way, multi-level analysis necessarily implies cultural specificity.

Cross cultural perspective. A short digression might be useful at this point. There is an assumption of much of the work in social science that the processes under study are relatively invariant across time and place. Triandis, rote this in their recent review of psychology and culture and it has long been a basic assumption of most of social psychology. The fundamental questions in cross cultural analysis, however, involve not only assumptions about similarities across cultures but differences across and within cultures.

Similarities across cultures have been termed etic phenomena by anthropologists. The fundamental problem is under what conditions does it make sense to derive principles that are truly applicable to different cultural situations. In other words, what principles can be abstracted which are not culture bound.

The corollary involves emic phenomena when the focus is on differences across cultures or on phenomena which are derived from and useful only in a specific cultural context. Emic concepts are especially germane to consumer behavior. The notions of market segmentation clearly recognize the necessity for understanding special sub-groups on their own terms. Another example is ethnomethodology which studies social phenomena by deriving meaning from the concepts and ideas which are relevant only to a particular culture or subset of that culture.

Thus, understanding social and consumer behavior involves not only general invariance across time and place but a delineation of the conditions of interaction with time and place. What is being suggested here is that a consideration of the concept of levels of analysis is necessary in identifying (1) when concepts and principles can be considered invariant across culture or other social units and (2) when principles need to be based primarily within a culture or considered as inevitably interacting with a time or place.



As seen in Figure 1, the interaction of time and the environment with levels of analysis is culture. Culture represents a multi-level analysis at a particular time and place. Because the levels of analysis are allowed to interact freely in the data of concern, culturally specific results are inevitable. If either time or the environment is held constant, and the other is varied, culture necessarily varies (culture 1 and 2 in Figure 1). If, however, a level of analysis is held constant across time or environment (represented by the horizontal bar in Figure 1), meaningful generalizations are possible since culture is not really invoked.

The relationship of the propositions presented in this paper to cultural behavior is basically that when cross cultural differences are analyzed within a level one may well develop principles which are cross cultural; when the concerns cross levels of analysis, however, specificity of culture is involved. When the analysis engages more than one level it is necessarily dealing with culturally specific concepts.

Proposition four suggests that the selection of a level of analysis dictates methodological and design questions. III the current pursuit of social and consumer science, there are clear methodological preferences and constraints depending on the level of behavior being studied. Certain problems of sociological analysis have used multivariate techniques, historical and archival data, or broadly aggregated interviews or questionnaires. Similar techniques have been the basis for broad marketing efforts Clearly the intention of such work is to describe a social system that represents some widely connected phenomena and has meaning in elaborating an aspect of social behavior. Methodological contrast is provided by social psychology and consumer decision processes. The laboratory has been widely used, not only because it represents a way of attacking relevant problems conceptualized at this level of analysis but also because it is consistent with the assumptions of this level (behavior is invariant across time and place)

Conceptualizing methodology in this way may also help alleviate the arguments about appropriate methods and the current concern over methodological clarity and generalization. The development of methods ought to be dictated within the confines of a level of analysis. Thus, laboratory work may remain the basic tool for consumer decision processes and individuals in the group research, so long as the variables tapped are relevant primarily to that level. Generalization from such research is constrained not so much by methodological considerations as theoretical considerations, i.e., what is the appropriate level of analysis for a particular problem.

Multivariate techniques may prove more useful within the confines of a level dealing with complex relationships among phenomena. When used across levels, multivariate and other broad quantitative methods are effectively culture bound (time and place) and may lead to no generalizable principles. Although generalizations across cultures may not be necessary, for example, in a marketing situation, this must be recognized by the researcher. Too often social scientists complain that complex methods yield uninterpretable and unreliable data. Concern for the collection of data by level should help clarify what data may or may not be relevant to the problem. The difficulty comes not with the amount of data generated by multivariate approaches but rather with the lack of clarity over whether or not data are meaningfully or spuriously related.

The fifth proposition suggests that each level generates its own theoretical statement. In other words, different processes may be used to understand behavior at different levels. This point argues that it may not be possible to integrate social theory if that means a social theory which simultaneously understands society, organizations, the group as a system, and individuals in the group. The lessons of physics suggest a metatheoretical approach which builds some indeterminacy into the social theoretical system. This would imply that social science is a set of theories, each explaining part of the variance depending on what level of social behavior is being analyzed and that analysis of one level precludes simultaneous integration of another level.

The sixth proposition is that a theoretical statement generated at a particular level may be unrelated to and even have paradoxical assumptions to explanations (theories) at other levels. For example, some of the traditional issues such as the concern over rationality-irrationality may not be issues at all. Rationality may be a relevant and important assumption at one level and irrationality at another. It may be appropriate to deal with man as irrational and groups as rational or vice versa.

Another example are the traditional concepts of utility and choice preference models. Recent research and theory (Tversky, 1972; Rapaport and Wallsten, 1972; Burns, 1972) suggests that utility needs to be greatly modified to serve as a basic approach to individual choice behavior. Nevertheless, it might be possible to develop a meaningful utility curve for organizations. Such paradoxes may have to be tolerated and even built into the study of social behavior.

The seventh idea observes that while the same behavior may be examined at different levels of analysis and that while behavior always involves people at different levels of analysis this is simply a confusing accident. Society is made up of individuals, organizations are made up of individuals, groups are made up of individuals. The observation that when we study societies, data from individual behavior is used and abstracted to that level is potentially confusing if the individual per se remains a focus. This may occasionally impede understanding society at the level at which it is appropriate. The observation that people in effect carry around society in their heads does not mean that one should simultaneously try to understand both the normative expectations of a large social system and personal style. Efforts to integrate in this fashion have provided grand social theory and little understanding of social behavior.

Proposition eight relates to earlier propositions in an effort to clarify problems between levels. Other levels, in which the level under analysis is obviously imbedded, have no necessary relevance for conceptualizing behavior at the level under investigation. For example, the individual in the group is part of a group as a system. In an effort to understand and deal with the group as a system there is no necessary reason for referring to or utilizing concepts relevant to either the individual in the group or organizations.

The ninth proposition states that there is no apriori basis on which to select a level as most appropriate or fruitful for social analysis. The concept of natural boundaries of phenomena suggests there is a direct relationship between the problem identified and the levels selected. An examination of consumer problems within or across cultures suggests a number of levels of analysis which could potentially and in fact have been fruitful in understanding consumer behavior. The consumption of goods or the problems of social behavior is an individual and systems problem and the implication here is that there may be no way of building a theory which bridges that gap.

Proposition ten suggests that understanding occurs by specifying relationships within a level. Functional statements and principles can be generated within a level that deal with that specific level and meaningfully integrate the set of phenomena bounded by that level. Theory and understanding or explanation are considered as specific to a level. Meaningful relationships across levels, from a theoretical point of view, may be impossible.

The final proposition is that the overall picture of theory in social behavior may require numerous levels, analyzed and portrayed independently with possible contradictions between levels but not within levels. The theory of social behavior is really a set of theories of social behavior, each subset accounting for some of the variance in a social situation but which necessarily cannot account for all of the variance at any one point in time.

In summary, these propositions present a context for the analysis of social and consumer behavior with special concern for what culture implies in the study of behavior. With the limiting conditions to generalization suggested here, it may well be that consumer behavior has more to offer the study of culture than cultural perspectives can offer to the understanding of consumer behavior. To understand consumer behavior at a time and place requires analysis of that time and place unless the analysis does not invoke more than one level. Since consumer behavior is persistent and pervasive, focusing on a level of analysis across time and place may well shed light on social and Personal behavior that is truly across cultures.


Berrien, K. F. General and Social Systems. New Brunswick: Rutgers University Press, 1968.

Burns, T. and Meeker, L. C. A mathematical model of multidimensional evaluation, decision-making and social interaction. In J. Cochrane and M. Zeleny (Eds.), Multiple Criteria Decision Making. Columbus: University of South Carolina Press, 1973.

Hansen, F. Consumer Choice Behavior. New York: Free Press, 1972.

Kuhn, T. Structure of Scientific Revolutions. Chicago: University of Chicago Press, 1962.

Rapaport, A. and Wallsten, T. Individual decision behavior. In P. Mussen and M. Rosengiveig (Eds.), Annual Review of Psychology, 1972, 23, 131-176.

Triandis, H., Malpass, R. and Davidson, A. Psychology and culture. In P. Mussen and M. Rosengiveig (Eds.), Annual Review of Psychology, 1972, 24, 355-378. Tversky, A. Elimination by aspects: A theory of choice. Psychological Review, 1972, 79, 281-299.



Gordon A. Haaland, University of New Hampshire


NA - Advances in Consumer Research Volume 01 | 1974

Share Proceeding

Featured papers

See More


The Effects of Being Time Poor and Time Rich on Happiness

Marissa Sharif, University of Pennsylvania, USA
Cassie Mogilner, University of California Los Angeles, USA
Hal Hershfield, University of California Los Angeles, USA

Read More


Changing the World, One Word at a Time: The Effect of Font Size on Prosocial Intention

Rima Touré-Tillery, Northwestern University, USA
Ayelet Fishbach, University of Chicago, USA

Read More


Associative versus Relational Processing: The Role of Elaboration in Evaluative Conditioning

Xiaomeng Fan, Northwestern University, USA
Galen V. Bodenhausen, Northwestern University, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.