ACR Fellow Speech

Richard P. Bagozzi,
[ to cite ]:
Richard P. Bagozzi (1994) ,"ACR Fellow Speech", in NA - Advances in Consumer Research Volume 21, eds. Chris T. Allen and Deborah Roedder John, Provo, UT : Association for Consumer Research, Pages: 8-11.

Advances in Consumer Research Volume 21, 1994      Pages 8-11

ACR FELLOW SPEECH

Richard P. Bagozzi

What I would like to talk about are three areas that have much potential for growth in consumer research in the years ahead. Before I do this, however, let me begin with a caveat. Since the inception of the ACR, the accomplishments I think we should be most proud of are those done by researchers in information processing, and to a slightly lesser extent by researchers in human judgment and choice. This is the heart of our field and the things I plan to talk about are not meant to detract from these accomplishments, but only to add to them. In fact, the information processing tradition in the ACR has produced fundamental contributions to basic knowledge and not only to substantive and applied consumer research. We need to continue this tradition. Another area we should take pride in is the general work done in qualitative research. This includes recent work in post modern aspects of consumption, ethnographic research, feminism, revival of interest in motivation research, and related areas. Our current and past editors of the JCR have done much in recent years to create an intellectual climate of diversity in the field. I am not aware of any field that has such a wealth of different ideas coexisting. Let us hope this continues, too. My comments to follow are not meant to take away from these developments. So let me turn now to three areas for future growth in the field.

When I mention the first area, many of you are likely to react with some skepticism. Nevertheless, I think there is a need for more research in the general area of personality in consumer research. By personality research, I'm not talking about the traditional theories and approaches. These were put to rest years ago by Walter Mischel's 1968 book and by Hal Kassarjian and colleagues in a series of articles.

By personality research, I mean research that is better grounded in individual differences, such as found in recent developments in the self concept and also in recent work in emotions in psychology. There has been a rebirth of sorts in personality research during the past few years which has resulted in better conceptualizations and better measurements that are only now starting to be looked at by consumer researchers. This work focuses on traits and states, as well as temperament.

When I say that there is room in the field for more research into personality, it is important to add that I am not particularly sanguine about personality, per se. Rather the real potential for the newer and better personality concepts lies not so much in their power as main effects or predictors but instead as moderating variables. Personality variables offer the most promise in conjunction with tests of hypotheses from the information processing tradition, for example. The best research using personality in consumer research has incorporated it as part of theory development and testing from other areas. For instance, the need for cognition, a scale developed a while back by Cacioppo and Petty, has recently been used by Curt Haugtvedt and colleagues to help explain the persistence of attitudes and the resistance of attitudes to change. Current research with the need for cognition shows that personality variables have promise when used to complement or qualify the effects of persuasive communication on information processing. Maybe we shouldn't be too surprised to find that consumer behavior can be explained by the interaction of individual difference and situational variables. The task for the future is to develop sound theories behind the interactions.

I used the need for cognition as an example of an individual difference variable that has been used successfully largely be- cause researchers have found a role for it within the context of a well-developed information processing theory; namely, the elaboration likelihood model. In its shorter version, the need for cognition is an 18-item scale. Actually, I think we are likely to see a movement away from the use of a large number of items to measure unidimensional personality concepts. The success of unidimensional scales with many items, such as the need for cognition, is likely to become a rarity. One reason is that researchers often do not have the luxury of being able to administer large scales within the context of their studies. Besides, why use 30 items when 8 to 10, say, might suffice? A second reason is that scales of 20 items or more often contain so much random error, and even systematic error, that it is unreasonable to expect unidimensionality.

In fact, more and more, we are finding that scales originally designed to reflect unidimensional constructs are multidimensional. Often the dimensions have been found to be more interpretable and to replicate better than the total scales. A good example of this is the self-monitoring scale which recently has been used in consumer research. The self-monitoring scale, a 25-item inventory, was originally devised to place people along a continuum from low to high self-monitors. The low self-monitoring individual guides his or her behavior on the basis of relevant inner resources such as values, feelings, attitudes, or dispositions. They are concerned that their own actions are accurate reflections of their underlying values and attitudes. By contrast, the high self-monitoring individual typically strives to be the type of person called for by each situation in which they find themselves. They are concerned about adjusting their actions to fit the interpersonal demands of the social situation. Internal states, such as attitudes, do not necessarily correspond to actions. Well recently, research has challenged this unidimensional representation. Instead, three distinct dimensions of self-monitoring have been found to underlie the scale. One of these is extraversion, a second is other directedness, and the third is acting ability. Researchers have found that these dimensions are not only correlated at low levels amongst themselves but two of them are even negatively correlated at times. Summing items across these dimensions obviously obscures the existence of distinct senses of self-monitoring. And in theory testing, because the dimensions frequently have distinct antecedents or lead to differential consequences, the use of the total scale can lead to misleading conclusions. Another scale used recently by consumer researchers, the self-consciousness scale, has also been found by psychometricians to be multidimensional. Researchers have found three distinct subcomponents: private self-consciousness, public self-consciousness, and social anxiety. We are likely to see greater attention paid to the subdimensions of these and other scales in the years ahead. We also need to re-examine past research based on the total summed scales to see if the conclusions drawn are really valid.

Very recently in personality research there has been a new movement that proposes a general hierarchical conceptualization of personality constructs. The framework proposes 4 alternative levels of abstraction. This movement has been led by the psychometrician Herbert Marsh and a few other psychologists (see also Bagozzi and Heatherton, 1994).

The first level in the hierarchy is the most abstract level or molar level of analysis and is, in fact, the traditional approach. Here all the items of a scale are added together to form a single construct. This has been labeled the total aggregation model in the literature, to indicate that an aggregation occurs across all items. In fact, if the scale is multidimensional, the aggregation will be across both items and dimensions. Needless to say this is a crude representation of personality constructs. The total aggregation approach is the one analyzed by Mischel, Kassarjian, and colleagues and found to be so lacking.

The other three levels in the hierarchy are less abstract and allow for multidimensionality. In the partial aggregation model we find a more molecular conceptualization of personality. Here each component of personality is represented as a separate indicator of a single higher order construct. Aggregation occurs across items but within dimensions. This point of view appears in current research in the JPSP on state self-esteem, among other constructs. The partial aggregation model has real advantages over the traditional total aggregation model with respect to the evaluation of reliability. Its major disadvantage is that it blurs the distinctions among multiple dimensions should these exist.

This is where the third conceptualization of personality in the hierarchical model comes to the fore. This approach has been termed the partial disaggregation model in the literature. Here each component or dimension of a construct is represented as a separate latent variable indicated by unique composites of subscales. In the past few months there has been an explosion of articles and papers using this approach in personality researchCmost of them in the psychology literature. It has been applied to self-esteem, self-monitoring, self-consciousness, and other personality traits and states. Some forthcoming research has used it in the modeling of emotions as well. The approach arose as a result of developments in psychometrics, but importantly, many researchers have found it especially useful in hypothesis testing and substantive research.

The fourth approach to the representation of personality constructs is labeled the total disaggregation model in the literature. This is the most atmostic of the four approaches. Here each individual item in a scale is modeled as a separate indicator of each dimension, where the items and dimensions are hierarchically organized. This approach has been applied with some success to the self-monitoring scale and to the need for cognition scale.

In sum, personality research has promise in consumer research but only if we use wellformed concepts and measures within the context of theory testing. One or more of the four levels in the inerarchical conceptualization of personality can be useful, depending on the researcher's purposes. There is a need to move away from the unidimensional scales and focus more closely on their subcomponents. But even this will only take us so far and is likely to be a transition to the next stage in personality research. Because most of the scales have been developed as unidimensional representations, it is likely that the subdimensions found empirically will be incomplete. The next phase in personality research should strive to better specify theoretically what the components are and then develop specific measures corresponding to them. Until we do this, the use of personality is likely to be a hit or miss affair in consumer research.

The second area for future growth in the field is the exploration of cross-cultural dynamics. This is an exciting, wide-open area. But we need to be careful. There is a tendency to take one theory developed in one culture and then apply it in another, searching for differences and commonalities. This approach helps us discover generalities and distinctions and might even lead to important serendipitous discoveries now and then. But it isn't enough and it isn't the way to build a foundation for research in the area. Culture isn't important unless we can specify what it is within and between different cultures that produces the commonalities and differences. Something is to be learned by taking existing theories and measures from one culture into another. But more progress will be made when we identify theoretical differences and develop measures valid across cultures. Many of our existing theories may generalize fully, or as a matter of degree, but there is a need to conceive of theories that explicitly address or incorporate cultural differences.

Cross-cultural research is possible for researchers from all points of view: information processing, human judgment and choice, attitude theory, and so on. To illustrate my points, I would like to take again the example of personality research, particularly the self concept. William James (1890) was an earlier theorist to define the self as an object of self-perception and self-knowledge (Smith, 1992). For him, the "empirical self or me" was the concept most central to personal experience. James maintained that the empirical self has three main divisions: the material self (i.e., one's body and possessions), the social self (i.e., the impression one conveys to others), and the spiritual self (i.e., one's inner or subjective being). Today the study of the self has a rich tradition in both psychology, where the self-concept is represented by such labels as self-schemas (e.g., Markus, 1977) and self-esteem (e.g., Fleming & Courtney, 1984), and sociology, where the self-concept is investigated as part of social-identity theory or role-identity theory, (e.g., Stryker, 1987).

We can think of the self-concept as a cognitive appraisal of the attributes about oneself (Hattie, 1992). The self-concept both mediates and regulates behavior. This is put nicely in the following quote from a recent article by Markus & Wurf in the Annual Review of Psychology: "The self-concept interprets and organizes self-relevant actions and experiences; it has motivational consequences, providing the incentives, standards, plans, rules, and scripts for behavior; and it adjusts in response to challenges from the social environment" (Markus & Wurf, 1987, pp. 299-230; see also Markus & Nurius, 1986).

For purposes of the talk today, I would like to draw on a number of articles by Triandis and especially a recent article by Markus & Kitayama in the Psychological Review. These researchers have found that two construals of the self can be identified in people, depending on the culture within which one has been raised (Markus & Kitayama, 1991; see also Triandis, 1989). The first, the independent self-concept is common in many Western cultures and is characterized by an emphasis on personal goals, personal achievement, and appreciation of one's differences from others. People with an independent self-concept tend to be individualistic, egocentric, autonomous, self-reliant, and self-contained. They place considerable importance on asserting the self and are driven by self-serving motives. The individual is the primary unit of consciousness, with the self coterminous with one's own body. Relationships with others frequently serve as standards of self-appraisal, and the independent self takes a strategic posture vis-a-vis others in an effort to express or assert one's internal attributes. One's personal attributes are primary and are seen as relatively stable from context to context. Emphasis is placed on displaying or showing one's attributes (e.g., pride, anger). The normative imperative is to become independent from others and discover one's uniqueness.

By contrast, the interdependent self-concept is common in many non-Western cultures and is characterized by stress on goals of a group to which one belongs, attention to fitting in with others, and appreciation of commonalities with others. People with an interdependent selfconcept tend to be obedient, sociocentric, holistic, connected, and relation oriented. They place much importance on social harmony and are driven by other-serving motives. The relationships one has are the primary unit of consciousness, with the self coterminous with either a group or the set of roles one has with individuals across multiple groups. Relationships with others are ends in and of themselves, and the interdependent self takes a stance vis-a-vis others of giving and receiving social support. One's personal attributes are secondary and are allowed to change as needed in response to situational demands. Emphasis is placed on controlling one's attributes (e.g., avoiding the display of anger). The normative imperative is to maintain one's interdependence with others and contribute to the welfare of the group.

The categories of independent and interdependent selves are, of course, ideal types and some variability is to be expected within any particular culture characterized by one or the other. Nevertheless, the distinctions are very real. Triandis (1989) describes how cultures shape either an independent or interdependent self. Markus and Kitayama (1991) argue that independent and interdependent selves have specific consequences for the acquisition and experience of cognition, emotion, and motivation. However, the ideas are so new that very little cross-cultural research can be found testing the existence of differences in the self-concept and their implications.

I would like to briefly mention one forthcoming study comparing the self-concepts of Japanese and American consumers (Abe, Bagozzi, and Sadarangani, 1994). The researchers began with the self-consciousness scale and found the following cultural differences. Japanese consumers had conceptualizations of the self that were more integrated and less distinct than Americans in the sense that self-images of private and public self-consciousness were more strongly associated in memory for them. The Japanese experienced higher levels of social anxiety than Americans, but lower levels of private self-consciousness. The researchers then related the dimensions of self-conscious to other personality scales recently used in consumer research. One of these was the attention to social comparison information scale, which is itself a refinement in the self-monitoring scale. The researchers found that attention to social comparison information was positively related to public self-consciousness and social anxiety, but unrelated to private self-consciousness. This was true for both Americans and Japanese. Although the authors did not point this out, these findings demonstrate my earlier point that components of personality scales often exist, in this case self-consciousness, and more importantly they function differently both within and across cultures. The researchers also investigated the action control scale, which also has been used by consumer researchers in the past. They found that action control was negatively related to public and private selfconscious-ness and social anxiety for both Americans and Japanese. In addition, action control was more strongly and negatively correlated with social anxiety for Japanese, as opposed to American, consumers.

Research like this is interesting, as far as it goes, in that it points out important boundary conditions for measures of the self-concept and its relationships to other constructs of interest both within and across cultures. The next step I hope the authors take is to show how these differences relate to decision making and choice in the marketplace. Cross-cultural research has potential for both providing a testing ground for our theories and suggesting refinements in our theories; maybe even leading to new theories.

The first two areas for future research I mentionedCpersonality and cross cultural researchCcomplement the core of our field and even provide avenues for enhancing the examination of the core. But they are likely to remain at the periphery. The third and final area for growth I wish to speak about is more fundamental. Moreover, there is opportunity for researchers in information processing and human judgment and choice to make contributions and push their ideas further in this area. The area I am referring to is the general notion of volition or will in consumer research.

For purposes of discussion it is useful to think about consumer action as taking place in two stages. The first might be labeled simply decision making. Here the reasons for acting are considered and a decision, plan, or intention to act is formed. By reasons for acting, I mean the consumer's assessment of product attributes and the consequences of consumption, the evaluation of these, comparisons of alternatives, and their integration in memory. This aspect of consumer behavior is, of course, where the most progress has been made in the field. A neglected aspect of decision making, however, is the study of how the reasons for acting get transformed or are translated into a commitment and plan to act. We need more research into the motivational processes leading up to volition. How exactly, and when, do judgments, attitudes, and preferences lead to intentions? We have left these processes unexplored but instead have assumed that intentions follow automatically upon attitude or preference formation. The missing mechanisms here seem to lie in particular affective and emotional processes and the coping responses the consumer works out in relation to them. More work is needed in this aspect of consumer decision making.

The issues I want to focus upon in the final minutes concern the processes following a decision or intention to act. We might label this aspect of consumer behavior self-regulation or action control (Bagozzi, 1992). Now for some products and services the translation of intentions into action is straightforward. We evaluate alternative brands, make a choice, and acquire the product immediately or almost immediately thereafter. The situations I want to address are those where the final act of consumption occurs after a gap in time and when impediments can thwart one's plans. Here future consumption in the mind of the consumer is problematic and, in fact, is a goal or objective to be reached. The question is how does the consumer enact his or her intentions? What are the processes of goal pursuit that lead to success or failure?

Well, it is useful to think of the processes as occurring in three stages: the consumer has to first initiate action, second sustain goal pursuit, and third terminate action. Let us look at the first stage. Here hesitations to act must be overcome. These reside largely in internal impediments such as fear of failure or a questioning whether the goal is really worth it. In this first stage appraisals are made about the means needed to achieve the goal. A decision maker assesses his or her self-efficacy with respect possible means to the end. In other words, the consumer evaluates how confidently he or she can perform the means. Judgments are made also about the likelihood that each means, if performed, would lead to goal attainment. And finally the emotional significances of the means are appraised. The initiation of goal pursuit will be a function of the integration of self-efficacy, instrumental beliefs, and affect toward alternative means.

Once the means are set into motion, the effort must be sustained if the goal is to be reached. In this second stage, the instrumental acts are monitored to see if they begin and end when they are supposed to do so, whether they achieve their objectives, whether new contingencies, impediments, or facilitating factors must be incorporated into decision making. Corrections to plans are made and commitment waxes or wanes, depending upon the situation, one's motivation, and judgments made in regard to progress toward the goal.

The final stage of goal pursuit is termination of goal pursuit. One issue here is when to abandon one's efforts when the goal is unattainable. Another is when do conditions require a change in goals. And even when a goal has been achieved, termination is sometimes an issue when the consumer cannot disengage from goal pursuit. We see this in the extreme in excessive dieting or over doing things in exercising or in general through various compulsive consumption behaviors. The subfield of customer satisfaction also addresses issues related to termination of goal pursuit that haven't been given much consideration to date.

The processes transforming an intention into goal attainment are obviously complex. No one study will be able to look at the entire sequence of decision making and goal pursuit. But many opportunities exist for investigating each of the subprocesses, and over time we should be able to gain a better picture of how volition functions in consumer decision making and action. This is an area where researchers from the information processing tradition can help shed light.

Let me give a brief example. Julius Kuhl, a German psychologist, has studied the activation of intentions (e.g., Kuhl, 1992). The probability that a given intention will be activated is a non-linear function of the importance, time pressure, and subjective competence associated with the intention. Kuhl found that people differ with respect to the processing of intention-related memory structures. State oriented individuals exhibit a highly generalized and inflexible tendency of decoupling perception and action that is accompanied by excessive reflections about past failures or future possibilities and alternative courses of action. This makes it difficult for state-oriented individuals to initiate new, non-automatized actions. By contrast, action-oriented people reveal processes that serve the recoupling of perception and action in order to realize a chosen intention. In his experiments, Kuhl found that action and state oriented individuals differed with respect to the activation of explicit intention representations in memory. In situations where the execution of intentions had to be postponed, action-oriented subjects deactivated declarative representations of prospective intentions when they were irrelevant to the current task. In contrast for state-oriented subjects, explicit representations of the postponed intentions persisted in a state of increased subthreshold activation. These results were found in a design based on recognition tasks and response time.

Many of the findings and the general experimental approach found in the information processing literature in consumer research are applicable to the study of volition. Information processing theory has much to offer concerning the formation, storage, retrieval, and activation of intentions. It can help us learn also about how goals are represented and integrated with decision making. The research on human judgment and choice can help in understanding how the means for goal pursuit are decided upon and how goal pursuit is guided or altered. These and many other questions remain to be explored. I believe the study of volitional processes needs to be added to the core of consumer research. Thank you very much.

REFERENCES

Abe, S., Bagozzi, R.P., & Sadarangani, P. (1994). An investigation of construct validity and generalizability of the self-concept: Self-consciousness in Japan and the United States. working paper, University of Michigan.

Bagozzi, R.P. (1992). The self-regulation of attitudes, intentions, and behavior. Social Psychology Quarterly, 55, 178-204.

Bagozzi, R.P., & Heathenon, T.F. (1994). A general approach to representing multifaceted personality constructs: Application to state self-esteem. Structural Equation Modeling, 1, 35-67.

Fleming, J.S., & Courtney, B.E. (1984). The dimensionality of self-esteem: II. Hierarchical facet model for revised measurement scales. Journal of Personality and Social Psychology, 46, 404-421.

Hattie, J. (1992). Self-Concept, Hillsdale, NJ: Erlbaum.

James, W. (1890). The Principles of Psychology (2 vols.). New York: Holt.

Kuhl, J. (1992). A theory of self-regulation: Action versus state orientation, self-discrimination, and some applications. Applied Psychology: An International Review, 41 (2), 97-129.

Markus, H. (1977). Self-schemas and processing information about the self. Journal of Personality and Social Psychology, 35, 63-78.

Markus, H.R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98,224-253.

Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41,954-969.

Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social psychological perspective. Annual Review of Psychology, 38, 299-337.

Smith, M.B. (1992). William James and the psychology of self. In M.E. Donnelly (Ed.), Reinterpreting the Legacy of William James (pp. 173-187). Washington, DC: American Psychological Association.

Stryker, S. (1987). Identity theory: Developments and extensions. In K. Yardley & T. Honess (Eds.), Self and identity: Psychosocial perspectives (pp. 83-103). New York: Wiley.

Triandis, H.C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96, 506-520.

----------------------------------------