Self-perception processes are most likely to occur when prior attitudes are

Although consumer behavior has been conceptualized as dynamic in nature, little empirical work has been done on the process by which behavior in one time period forms the basis for subsequent attitudes and actions. Recent developments in self-perception theory, however, provide the analytical tools necessary for investigating this process. The basic postulates of and evidence for self-perception are reviewed, and current research issues are identified and discussed in this paper. Implications for consumer research are then considered.

INTRODUCTION

Consumer behavior has been conceptualized traditionally as a dynamic process in which actions are affected by what is learned from previous behavior as well as by in- formation from external sources (e.g., advertisements, friends, relatives, etc.). This property of dynamism is incorporated implicitly in most research (e.g., studies which control for past experience by using new or fictitious brand names), and explicitly in theoretical statements which include feedback loops from behavioral to cognitive variables (e.g., Engel, Kollat & Blackwell, 1973; Howard & Sketch, 1969; Nicosia, 1966) and in research utilizing Markov, linear learning, or stochastic models of brand choice (e.g., Bass, 1974; Kuehn, 1962). While this link between past and future behavior is acknowledged, little research attention has been focused on the process by which behavior in one time period forms the basis for subsequent attitudes and actions.

Recent theoretical developments provide the analytical tools necessary for exploring this rich and exciting area. Specifically, self-perception theory (Bem, 1965, 1972) attempts to explicate how an individual interprets his own behavior, how he assigns meaning to that behavior, and under what conditions the individual accepts experiential information as valid and worthy of incorporation into his attitude and behavior set. Self-perception theory's central province is the feedback loop, often postulated by theory and often neglected empirically.

Although self-perception theory has been developed within social psychology, it has major implications for the study of consumer behavior, and this is the primary focus of this paper. First, the basic postulates of self-perception theory are reviewed. Next, basic evidence for the theory is presented. Several current issues of relevance to consumer researchers are identified and research related to them are discussed. Finally, the implications of this work for consumer behavior theory and research are considered.

SELF-PERCEPTION THEORY

Self-perception theory (Bem, 1965, 1972) can be viewed as one portion of a general set of propositions commonly referred to as attribution theory. Like other attributional paradigms (see Kelley, 1967 on object attribution, Jones & Davis, 1965 on other person attribution, and Kelley, 1973 for an integrated view of all three paradigms), self-perception theory is phenomenological in character. That is, self-perception theory takes the perspective of the actor. It attempts to explain and predict how individuals come to understand the causes of their own behavior, and to specify the consequences of this causal assignment.

The basic postulate of self-perception theory is that:

Individuals come to "know" their own attitudes, emotions, and other internal states partially by inferring them from observations of their own overt behavior and/or the circumstances in which this behavior occurs (Bem, 1972, p. 2). [A second postulate of self-perception theory is that individuals are functionally in the same position as an outside observer who must rely on external cues to infer the individual's inner state (Bem, 1972). While this proposition has caused a substantial amount of controversy because it implies that individuals have no more information about their feelings than observers, it has been shown to be tenuous at best (see Jones and Nisbett, 1971). Negation of this second postulate, of course, does not invalidate the first one which has received widespread support.]

Self-perception theory postulates the mechanism and the conditions under which a person uses his own behavior as data to make inferences about himself. For behavior to be used to infer individual characteristics, it must be a credible indicator of his internal state. In order to be credible, it must be perceived by the person as resulting from his internal motivations or true reactions to a stimulus, i.e., it must be attributed to properties of oneself as opposed to controlling aspects of the environment. [A note on terminology should be made here. Many authors refer to this process as internal (self) versus external (situational contingency) attribution, while others (notably Nisbett and Valins, 1971) prefer the terms stimulus (perceived intrinsic properties) versus circumstance (factors external to the stimulus). The difference is more apparent than real. Behavior that is perceived to be elicited by the person's intrinsic feelings about the stimulus (stimulus attribution) implies internal motivation or causality. Behavior elicited by a desire to gain a particular reinforcement or to avoid punishment or by some other aspect of the circumstances (circumstance attribution) implies external causality. For simplicity, the terms "internal" and "external" will be used here.] Thus, perceived causality of behavior is a primary determinant of the belief inference process.

Perception of causality, as indicated by Bem's proposition, depends partially upon the circumstances in which behavior is enacted. According to the discounting principle (Kelley, 1971), "the role of a given cause in producing a given effect is discounted if other plausible causes are present" (p. 8). [Circumstantial factors can enhance the probability of self-attribution. Behavior enacted under inhibitory conditions (e.g., purchase at a premium price) should enhance the perception of internal motivation and result in strong belief inferences (augmenting principle, Kelley, 1971). These factors have been investigated only rarely, and thus are not discussed here.] The presence of plausible external causal factors such as reward/cost contingencies, lack of volition, and the like should induce discounting of internal motivations as the cause of behavior. If a person performs an act for a large reward, for example, he would attribute his behavior to the reward rather than to his own positive attitude toward the act. When attributions are made to such external situational contingencies ("I was paid to do this"), no belief inferences would be made, and no link between behavior in one situation and behavior in another is expected. On the other hand, attributions to oneself ("I did this because I wanted to") are postulated to result in belief inferences ("I must believe in this issue") and subsequent behavior consistent with the initial behavior and internal attributions.

This attribution process is diagrammed in Figure 1. The relationship between past behavior, discounting cues, and future attitudes and actions has received the most empirical attention, and this body of literature provides the strongest support for the theory. The other parts of the process have received less attention, and these relationships are the source of current issues in self-perception research. The next section provides brief overview of this work.

FIGURE 1

A SCHEMATIC OF THE ATTRIBUTION PROCESS

EMPIRICAL EVIDENCE

Evidence relevant to self-perception theory can be discussed by first examining those studies which focus on the effects of past behavior, and then examining those that deal with the effects of perceived causality of behavior derived from Bem's (1972) central proposition. Several questions regarding the self-perception process emerging from these basic works and preliminary findings pertinent to them can then be discussed.

Effects of Past Behavior

At a minimum, two conditions must be demonstrated if the basic postulate of self-perception theory is valid. First, it must be shown that individuals do use their own behavior to infer their attitudes and/or to guide subsequent behavior. Second, it must be shown that circumstances surrounding the enactment of behavior affect this inference process.

The first of these conditions has been illustrated by several laboratory and field experiments in which a behavior is induced and subsequent attitudes and/or behavior are measured (e.g., Bandler, Madaras, & Bem, 1968; Freedman & Fraser, 1966; Pliner, Hart, Kohl, & Saari, 1974; Snyder & Cunningham, 1975; Valins, 1966). In a laboratory study (Valins, 1966), for example, male subjects were allowed to hear what was ostensibly their own heart beat while viewing pictures of nude females. Pictures associated with a (falsely manipulated) quickened heart rate were subsequently rated as more attractive and were frequently chosen as ones the subjects would like to keep than others. Similarly, in a field study utilizing what has become known as the foot-in-the-door paradigm, people were contacted at home and asked to display a small sign promoting a social cause in their window (Freedman & Fraser, 1966). Two weeks later, these individuals and a control group of subjects not contacted previously were asked to comply with another very large request. When both the issue and the task were similar across requests, 76% of the experimental participants as opposed to 16.7% of the control participants agreed to comply with the large request. Apparently, the initial behavior was self-attributed, causing subjects to draw belief inferences which enhanced the likelihood of subsequent compliance.

Effects of Perceived Causality

The second necessary pattern of data has been reported in numerous experiments which show that behavior enacted under discounting cue conditions does not persist in future time periods and may even decline in frequency (Bem, 1966; Davison & Valins, 1969; Calder & Staw, 1975; Deci, 1971; Kruglanski, Freedman, & Zeevi, 1971; Kruglanski, Alon, & Lewis, 1972; Kruglanski, Riter, Arati, Agassi, Monteqio, Peri, & Peretz, 1975; Lepper, Greene, & Nisbett, 1973; Lepper & Green, 1975; Storms & Nisbett, 1970). Using an overjustification paradigm, [Overjustification situations are the converse of insufficient justification ones, are those in which there are both internal and external motivations for engaging in an activity.] for example, Lepper et al. (1973) found that children who expected and received a reward for engaging in a playtime activity subsequently spent less time engaging in that activity than a control group of children, even though they had previously evidenced considerable interest in it.

More recently, both of the two necessary conditions have been demonstrated in the consumer behavior literature. Scott (1976) asked individuals to take a two-week trial subscription to a weekly community newspaper either at the regular price (discounting cue absent) or at various discounted prices (discounting cues present). After the trial period had expired, these participants and a control group of subjects not receiving the trial offer were called and asked to subscribe on a regular basis. The results supported self-perception predictions in that large incentives for trial produced no more regular subscriptions than did the no-trial control condition. No or very small incentives, however, did have a positive effect on subscription behavior.

The negative effect of monetary incentives on subsequent behavior has been replicated in a social marketing context (Scott, in press) and in an interview request situation (Reingen & Kernan, 1977). And, Tybout (in press) and Dholakia & Sternthal (1977), using essentially the same foot-in-the-door paradigm, have shown that high credibility of the source of a message or request also acts as a discounting cue which attenuates subsequent behavior. Finally, Hansen (in review) has found that while monetary inducements to respond to a mail survey do not adversely affect response rate, they do result in lower quality of responses.

These studies provide considerable support for the self-perception thesis, and demonstrate the applicability of the self-perception framework for many consumer behavior contexts. However, they and other experiments conducted more recently raise some fundamental questions regarding the self-perception process and indicate areas where theoretical extensions and empirical research are needed to clarify our understanding of the link between past and future behavior. Six of these issues and preliminary data relevant to them will he discussed next.

RESEARCH ISSUES

Many areas of the self-perception process remain relatively unexplored. Following the diagram in Figure 1, these areas may be categorized as relating to the strength of the behavior-behavior link, the definition of past behavior, the relationship between the magnitude of discounting cues and behavior, the mediating role of attributions, the outcomes of self-perception processes, and the way in which experiential information is combined with other types of information. Research issues in each area and their implications will be briefly considered.

Issue 1: Strength of Self-Perception Effects

A very basic issue relates to the strength of the link between past and future actions. In their seminal study; Freedman and Fraser (1966) suggest that an initial behavior has a rather dramatic and persistent (two weeks) effect on subsequent actions. However, only behavioral intentions to comply with the large request were measured. Actual behavior was measured by Pliner et al. (1974), but the requests were made only one day apart, and difficulty of the behavior (e.g., donation to charity) was a dependent variable rather than an independent one. Scott (in press) varied second request task difficulty and measured both behavioral intentions and behavior in order to assess directly the strength of the initial behavior effect. The findings replicated those of Freedman and Fraser (1966) in that behavioral intentions to comply with the very difficult task were positively affected, but behavior was affected only when the second task was moderately difficult.

These studies show that an initial behavior does have a positive influence on subsequent behavior, but that this effect may be limited by the difficulty of that subsequent action. It seems likely that a series of requests could be used to gain compliance with very large requests, but this speculation has not been tested empirically. Thus, the strength of influence strategies based on this paradigm is still unclear at least in the case where the two requests are made on two different occasions and where behavioral intentions are not synonymous with behavior. Other studies have examined the effects of size of the initial behavior on subsequent behavior, and these are discussed next.

Issue 2: What is Past Behavior?

In the field studies discussed previously (e.g., Freedman & Fraser, 1966; Pliner et al., 1974; Scott, 1976, in press), the initial behavior consisted of behavioral intentions to comply with a request. Two other types of experiments have varied this definition of past behavior. The first examines the effect of merely relabeling or reinterpreting some behavior that has already occurred, while the other focuses on enhancing the behavior-behavior link by increasing the level of difficulty of an initial, induced behavior.

Labeling studies manipulations are often no more than the provision of a belief-relevant label or interpretation for behavior that has already occurred. In a study by Tybout and Yalch (in review), for example, individuals responded to a survey about voting behavior. Half the individuals were then told that they were more likely to vote than the average person, while others were told that they were about average in their likelihood of voting. Examining records of a local election one week later, the authors found that more individuals given the "above average" feedback had indeed voted than individuals given the "average" feedback. These findings replicate those of Kraut (1973) who found that labeling individuals as charitable or uncharitable re-suited in subsequent behavior consistent with these labels, and of Miller, Brickman, and Bolen (1975) who showed that labeling children as non-litterers or math achievers enhanced subsequent behavior relevant to that label. All of these results are consistent with self-perception assumptions about the cognitive status of attitudes and empirical evidence which indicates that people often do not know what their attitudes are (e.g., Ross, Insko, & Ross, 1971). They are also congenial with findings that attitudes are inferred from behavior only when behavior is perceived as belief-relevant, or belief expressive (Kiesler, Nisbett, & Zanna, 1969). That is, labeling affects behavior because it implies a link between past actions and beliefs, and makes this link salient to the individual.

All of these studies, however, have utilized a personal contact to deliver the label. Extending this work, Allen (1977) tested the relative effectiveness of a labeling message versus a typical persuasive message using television advertisements. The results were equivocal, but suggest that under some conditions people are sensitive to the explicit positive label ("You are the kind of person who...") of the labeling message and the implicit negative label of the persuasive message ("You ought to be..."). Too little is known about the qualifying conditions of labeling effects and the cognitive meaning of labels to reach definite conclusions, but the question of the situations in which labeling or persuasive messages are most effective is an intriguing one for consumer researchers.

A second type of study takes an approach opposite that of labeling, and seeks to test methods of strengthening the effect of an initial behavior by increasing its difficulty or size. Seligman, Bush, and Kirsch (1976), for example, found that very small behaviors had no effect on subsequent actions, while Pliner et al. (1974) found no differences between groups performing different levels of initial behaviors. It is clear from even a cursory examination of these discrepant findings that we actually know very little about what kind of an initial behavior is required for the self-perception process to become operative, nor do we know along what dimensions this initial behavior should be classified. It may be that size of the request is not as important as degree of involvement, for example. These questions must be answered before influence strategies predicated on the paradigm can be used confidently.

Issue 3: Magnitude of Discounting Cues and Behavior

The definition of discounting cues as well as the definition of past behavior is an issue of particular relevance for consumer research. In many previous studies, a discounting cue is either present or absent. In contrast, consumer choice contexts often involve an initial behavior or trial of a product induced by a monetary incentive (e.g., coupons, cents-off deals, etc.) which may be of a variety of sizes. While self-perception theory does not specifically hypothesize a relationship between magnitude of the discounting cue and the likelihood of subsequent behavior, a reasonable derivation is that this relationship should be negative. Scott (1976; in press) addressed this issue empirically by including several levels of monetary incentives as treatments. Two findings are of interest. First, small incentives did not produce discounting in a commercial setting (Scott, 1976). Second, the relationship between magnitude of the incentive and subsequent behavior appeared to be negative, but not linear, in both a commercial (Scott, 1976) and a social issue setting (Scott, in press).

The first of these findings suggests that monetary incentives as discounting cues may have different effects in commercial than social settings. In commercial contexts, small incentives for trial may be perceived as so intrinsic to behavior that they are not perceived as salient causes of behavior. Kruglanski (1975) has shown that rewards intrinsic to the task (e.g., stock market choices) do not result in decreased interest or behavior as they do when the rewards are not intrinsically related to the task (e.g., a social game). Thus it may be that small monetary incentives do not lead to discounting of internal motivations in commercial contexts. It is only when they become very large, or when they are used in an unusual setting (e.g., a social program), that they are perceived as salient causes of behavior and discounting occurs. The findings of a single experiment should, of course, be regarded as the only tentative evidence at best. However, researchers need to be sensitive to areas where cross-situation-al generalizability may be problematic.

The second finding indicates that incentives as discounting cues may carry different meanings and thus activate different processes than other forms of discounting cues. That is, they not only provide contextual cues about behavior, but also specify the terms of an exchange in which perceptions of equity (Adams 1965) are relevant. When incentives are very large relative to the task, the person may feel that he has been inequitably rewarded and is obligated to do something in return to restore equity. An attempt to investigate this issue would be characterized by an experiment including several levels of incentives and measures of attitudes and perceptions of inequity as well as behavior. If the speculations about equity are correct, behavior would be curvilinearly related to levels of incentive, but attitudes would be negatively related and perceptions of inequity positively related to the magnitude of incentives. This type of research on the meaning and effects of monetary incentives is of central importance in understanding much marketplace behavior.

Issue 4: The Mediating Role of Attributions

Many self-perception studies merely assume the presence of an attributional mechanism from behavioral findings consistent with attribution predictions. However, varying the presumed antecedents of self-perceptions and observing the outcomes is not sufficient to demonstrate the efficacy of the theory (see Burnkrant, 1975). Direct assessment of attribution is necessary since competing theories often make the same behavioral predictions.

Among studies which do include measures of cognitive processes, a frequent finding is significant behavioral effects, but no or only slight effects on cognitive variables assumed to mediate behavior (e.g., self-perceptions, attitudes, attributions). An interesting theoretical rationale for these results is one proposed by Nisbett and Valins (1971) who postulate that self-attributions and the corresponding self-perceptions or beliefs have a cognitive status similar to an hypothesis. That is, performance of a behavior causes individuals to attribute it to themselves, and this attribution in turn causes the individual to form an hypothesis about his beliefs, which is then tested out in subsequent actions. The behavioral effects demonstrated in previous studies then result from a desire to test this attributional hypothesis by gaining more information rather than to act consistent with stable attributions and attitudes.

The hypothesis-testing proposition is compatible with the results of several studies. For example, Barefoot and Straub (1971) were able to replicate Valin's (1966) false heart rate feedback findings only when subjects were given sufficient time to examine the pictures. Apparently the quickened heart rate caused subjects to view the pictures more closely and confirm that they were indeed more attractive. Similarly, Valins (1974) and Ross, Lepper, and Hubbard (1975) have demonstrated that merely debriefing subjects about the falsity of the feedback had no effect. Again, it seems that feedback stimulates an individual to gather more information to confirm the belief hypothesis. Once confirmed, the falsity of the feedback is irrelevant to the belief.

A recent experiment tested this attribution process directly (Scott & Yalch, in press). Subjects were asked to test a new soft drink under incentive or no incentive conditions. After making this choice, half the subjects were allowed to visually examine the product. All subjects then tasted the product which was manipulated to provide either a good, neutral, or poor experience. When subjects were allowed to examine the product, the familiar discounting results were found in the neutral and poor taste conditions; incentives undermined subjects' attitudes toward the drink, behavioral intentions to purchase it, and actual choice of it relative to no-incentive conditions. The reverse effect of incentive was found when no examination opportunity was given. It appears that individuals do infer their beliefs from their own behavior, but only when they have an opportunity to persuade themselves of the veracity of the experimental information.

The hypothesis-testing proposition and related empirical work represent beginning attempts to outline the conditions under which attributional hypotheses are translated into beliefs and behavior. Unfortunately, this begs the question of what starts the attribution process in the first place. That individuals do not always engage in causal analyses is suggested by findings from Tybout (in press). In her study, the predicted discounting cue effects were obtained only when subjects' attention was drawn to their behavior by a question asking why they behaved as they did.

The issue is not really whether individuals do or do not engage in attributional analyses, but rather is "under what conditions do they perform these analyses." The answer may be found in terms of individual difference characteristics (e.g., absence of self-schema on a dimension as in Markus (1977), internal versus external locus of control as in Shaver (1976) or low education as in Tybout, in press). Or, it may be found in terms of the characteristics of the context of behavior (e.g., expectation of future consequences as suggested by Berscheid, Graziana, Monson, & Denmer, 1976, and relevance of the behavior to beliefs as in Kiesler, Nisbett, & Zanna, 1969). Unfortunately, little evidence exists to answer this question conclusively although the literature certainly provides some hypotheses. It is critical that future research address this issue since self-perception theory will have little applicability to consumer behavior if it turns out that only a small segment of the population engage in these analyses spontaneously and that there are no methods to encourage such analysis. In sum, the external validity of the self-perceptions process must be defined and demonstrated.

Issue 5: The Cognitive Consequences of Self-Attributions

Related to the issue of when attributions are made and their cognitive status once they are made is the issue of what happens as a result of self-attributions. Much evidence exists which demonstrates a behavior change, but we know very little about the cognitive changes that occur. In interpreting their results, Freedman and Fraser (1966) offered the following possibilities:

Once he has agreed to a request, his attitude may change. He may become in his own eyes, the kind of person who does this sort of thing, who agrees to requests made by strangers, and who takes action on things he believes in, who cooperates with good causes (p. 201).

Thus behavior may change a person's perception of himself in general, or may alter attitudes specifically related to the behavior, or both. And, there is some evidence to suggest that other cognitive activities occur as well.

With regard to the self-perception versus specific attitude question, a few studies indicate that one's self-perception is most affected. Freedman and Fraser (1966), for example, reported positive (but not significant) effects of past behavior even when the issue and task were different across requests. Similarly, significant positive effects of behavior on subsequent actions have been observed when the issue was different across requests (Snyder & Cunningham, 1975). These studies provide indirect evidence that issue-specific attitude change is not solely responsible for the behavioral effects.

Direct evidence for an effect on self-perception is found in Scott (in press). Attitudes and perceived personal activism were both measured in this study, with larger (although not significant) effects found on activism than on issue-specific attitudes. Some methodological problems exist in this design, however, and the results must be viewed as tentative. And, attitudes of the subject population were already favorable, so a ceiling effect may have been encountered.

Future research is needed which clarifies the cognitive effects of self-attributions since this will be a critical factor in determining the applicability of self-perception based strategies to consumer settings. If issue-specific attitudes are not affected, the applicability of the theory to choice decisions among similar alternatives is problematic. On the other hand, there are also many situations in which attitudes are already favorable but behavioral enactment of these attitudes is low (e.g., everyone believes that conservation of energy is important, but few take action), and changes in self--perceptions may provide an energizing force on behavior.

Finally, there is evidence to suggest that the attribution process may affect other cognitive variables indirectly. Several studies show that providing feedback about one's behavior, or making that behavior salient instigates a search for information (e.g., Taylor, 1975; Berscheid et al., 1976). And, other experiments indicate that the provision of rewards or incentives for behavior also have this effect. Using a Bayesian framework for analyzing the value of information, Yalch and Scott (in review) and Scott and Tybout (1977) have found that subjects who agreed to test a product under incentive (reward) conditions responded more to information provided them after this choice.

Issue 6: Combining Experiential and Other Types of Information

Self-perception theory offers fairly clear specifications of when an individual uses his own behavior to formulate attitudes and guide behavior. Other attributional paradigms (e.g., object and other person) suggest other types of information that may be used to form judgments. Specifically, distinctiveness of response, consistency of response, and consensus information may be used when the person forms judgments over time (see Kelley, 1976; 1973). It seems likely in most consumer contexts that individuals will have access to at least some of these types of information as well as to information from their own experiences. We know that self-perception induced beliefs can affect both the quantity of information and the perception of the content of that information at least when it is somewhat ambiguous (e.g., tasting a "new" soft drink as in Scott & Yalch, in press). The question now is whether these beliefs will be given greater or lesser weight than information of another type.

Burnkrant and Cousineau (1975) have shown that individuals' judgments of product quality are influenced by the degree of consensus exhibited by others. The greater the apparent consensus, the more an individual is influenced by this information. These authors did not, however, examine the relative influence of experiential and consensus information. That is, they did not induce a self-perception based belief and observe changes in those beliefs as a function of consensus information. Research by Nisbett, Borgida, Crandell, and Reed (1976) indicates that individuals often ignore consensus information and rely more heavily on their own experiential data, and Scott and Tybout (1977) found that the timing of consensus information presentation as well as whether discounting cues are present affect the acceptance of consensus information. That is, rewarded subjects showed a primacy effect in that they were more influenced by consensus information when it preceded actual product experience. Unrewarded subjects generally were less responsive to both consensus and actual product experience.

The process by which information sources are combined is undoubtedly a complex one which will require much research to unravel. It seems clear, however, that a merging of the paradigms is necessary to extend the external validity of the research questions posed and the experimental designs used in consumer behavior research.

IMPLICATIONS AND CONCLUSIONS

Self-perception theory and research promises to add much to our understanding of how consumers learn from their own experiences and the consequences of this learning for future actions. The phenomenological focus of this framework forces one to examine not only behavior itself, but also to examine how the individual perceives the behavior. The meaning an individual assigns to his experiences is of critical importance in predicting and explaining subsequent actions. Self-perception theory acknowledges that learning from past behavior is not a simple, automatic response. Rather, it is a complex process in which what is learned from previous experiences is dependent upon many interacting factors. While situational factors affect the perception of the outcomes of consumer behavior, for example, it is also likely that some outcomes will affect the perception of situational cues.

Self-perception theory (and attribution theory in general) is inherently dynamic, and thus it orients researchers to examine behavior over time. Situational characteristics, in self-perception terms, affect behavior in both immediate and future time periods. Investigation of only the immediate impact of situational variables tells only half the story. Similarly, strategies designed to modify behavior must be evaluated at several points in time since their long-run effects may vary from their short term ones.

Caution must be exercised, however, in expecting too much of self-perception theory as developed thus far. Consumer researchers must be aware of the unresolved issues surrounding the self-perception process, and must be vigilant for conditions which may limit its cross-situational generalizability. As noted in this paper, we have only just begun to examine the self-perception process in any depth. And, consumer researchers must be careful even in applying the knowledge gained in other contexts. In consumer contexts, it may be that some product decisions involve attributional processes more frequently than others, that some individuals utilize attributional analyses more than others, or that special strategies must be developed to encourage attributional processes.

It should be clear from this paper that much research remains to be done in explicating the self-perception process and its implications for consumer behavior. Given its potential for shedding light on that ubiquitous feedback loop, however, such research seems well worth our time and resources.

REFERENCES

J. Stacy Adams, "Inequity in Social Exchange," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 2 (New York: Academic Press, 1965), 267-299.

Chris T. Allen, "Delimiting and Describing the Phenomenon of Socially-Conscious Consumption Behavior: An Experimental Approach," Unpublished dissertation, The Ohio State University, 1977.

R. G. Bandler, G. R. Madaras, and D. J. Bem, "Self-Observation as a Source of Pain Perception," Journal of Personality and Social Psychology, 9 (July 1968), 205-209.

J. C. Barefoot and R. B. Straub, "Opportunity for Information Search and the Effect of False Heart-Rate Feedback," Journal of Personality and Social Psychology, 17 (February 1971), 154-157.

Frank M. Bass, "The Theory of Stochastic Preference and Brand Switching," Journal of Marketing Research, 11 (February, 1974), 1-20.

D. J. Bem, "An Experimental Analysis of Self-Persuasion," Journal of Experimental Social Psychology, 1 (August, 1965), 199-218.

D. J. Bem, "Self-Perception Theory," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 6 (New York: Academic Press, 1972), 1-62.

D. J. Bem, "Inducing Belief in False Confessions," Journal of Personality and Social Psychology, 3 (June, 1966), 707-710.

Ellen Berscheid, William Graziano, Thomas Monson, and Marshall Dermer, "Outcome Dependency: Attention, Attribution, and Attraction," Journal of Personality and Social Psychology, 34 (November, 1976), 978-989.

Robert E. Burnkrant, "Attribution Theory in Marketing Research: Problems and Prospects," in M. J. Schlinger (ed.), Advances in Consumer Research, Vol. 2 (Association for Consumer Research, 1975), 465-469.

Robert E. Burnkrant and Alain Cousineau, "Informational and Normative Social Influence in Buyer Behavior," Journal of Consumer Research, 2 (December, 1975), 206-215.

B. J. Calder and B. M. Staw, "Self-Perception of Intrinsic and Extrinsic Motivation," Journal of Personality and Social Psychology, 31 (April, 1975), 599-605.

G. Davison and S. Valins, "Maintenance of Self-Attributed and Drug-Attributed Behavior Change," Journal of Personality and Social Psychology, 11 (January, 1969), 25-33,

E. L. Deci, "The Effects of Externally Mediated Rewards on Intrinsic Motivation," Journal of Personality and Social Psychology, 18 (April, 1971), 105-115.

Ruby Roy Dholakia and Brian Sternthal, "Highly Credible Sources: Persuasive Facilitators or Persuasive Liabilities?'' Journal of Consumer Research, 3 (March, 1977), 223-232.

James F. Engel, David T. Kollat, and Roger D. Blackwell, Consumer Behavior, (New York: Holt, Rinehart, and Winston, 1973).

J. L. Freedman and S. C. Fraser, "Compliance Without Pressure: The Foot-in-the-Door Technique," Journal of Personality and Social Psychology, 4 (August, 1966), 195-202.

Robert A. Hansen, "The Effects of Monetary and Nonmonetary Incentives on Mail Survey Respondent Behavior: A Self-Perception Interpretation," in review.

John A. Howard and Jagdish N. Sheth, The Theory of Buyer Behavior (New York: John Wiley, 1969).

E. E. Jones and K. E. Davis, "From Acts to Dispositions," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 2 (New York: Academic Press, 1965), 219-266.

H. H. Kelley, "Attribution Theory in Social Psychology," in D. Levine (ed.), Nebraska Symposium on Motivation (Lincoln, Nebraska: University of Nebraska Press, 1967).

H. H. Kelley, "Attribution in Social Interaction," in E. E. Jones et al., Attribution: Perceiving the Causes of Behavior (Morristown, New Jersey: General Learning Press, 1971), 151-173.

H. H. Kelley, "The Processes of Causal Attribution," American Psychologist, 38 (February 1973), 107-128.

C. A. Kiesler, R. E. Nisbett, and M. Zanna, "On Inferring One's Beliefs From One's Behavior," Journal of Personality and Social Psychology, 11 (April, 1969), 321-327.

R. E. Kraut, "Effects of Social Labeling on Giving to Charity," Journal of Experimental Social Psychology, 9 (November, 1973), 551-562.

A. W. Kruglanski, "The Endogenous-Exogenous Partition in Attribution Theory," Psychological Review, 82 (November, 1975), 387-406.

A. W. Kruglanski, S. Alon, T. Lewis, "Retrospective Misattribution and Task Enjoyment," Journal of Experimental Social Psychology, 8 (November, 1972) 493-501.

A. W. Kruglanski, I. Freedman, and G. Zeevi, "The Effect of Extrinsic Incentive on Some Qualitative Aspects of Task Performance," Journal of Personality, 39 (December, 1971), 606-617.

A. W. Kruglanski, A. Riter, D. Arazi, R. Agassi, J. Monteqio, I. Peri, and M. Peretz, "Effects of Task Intrinsic Rewards Upon Extrinsic and Intrinsic Motivation,'' Journal of Personality and Social Psychology, 3 (April, 1975), 699-705.

A. A. Kuehn, "Consumer Brand Choice as a Learning Process,'' Journal of Advertising Research, 2 (December, 1962), 10-17.

M. R. Lepper and D. Greene, "Turning Play into Work: Effects of Adult Surveillance and Extrinsic Rewards on Children's Intrinsic Motivation," Journal of Personality and Social Psychology, 31 (March, 1975), 479-486.

M. R. Lepper, D. Greene, and R. E. Nisbett, "Undermining Children's Intrinsic Interest With Extrinsic Reward: A Test of the Overjustification Hypothesis," Journal of Personality and Social Psychology, 28 (October, 1973), 129-137.

Hazel Markus, "Self Schemata and Processing Information About the Self," Journal of Personality and Social Psychology, 35 (February, 1977), 63-78.

R. L. Miller, P. Brickman, and D. Bolen, "Attribution Versus Persuasion as a Means of Modifying Behavior," Journal of Personality and Social Psychology, 31 (March, 1975), 430-41.

Francesco M. Nicosia, Consumer Decision Processes (Englewood Cliffs, New Jersey: Prentice-Hall, 1966).

R. E. Nisbett, E. Borgida, R. Crandall, and H. Reed, "Popular Induction: Information is not Always Informative,'' in J. Carroll and J. Payne (eds.), Cognition and Social Behavior (Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1976).

R. E. Nisbett and S. Valins, "Perceiving the Causes of One's Own Behavior," in E. E. Jones et al., Attribution: Perceiving the Causes of Behavior (Morristown, New Jersey: General Learning Press, 1971), 63-76.

P. Pliner, H. Hart, J. Kohl, and D. Saari, "Compliance Without Pressure: Some Further Data on the Foot-in-the-Door Technique," Journal of Experimental Social Psychology, 10 (January, 1974), 17-22.

P. H. Reingen and J. B. Kernan, "Compliance With an Interview Request: A Foot-in-the-Door, Self-Perception Interpretation," Journal of Marketing Research, 14 (August, 1977), 365-369.

M. Ross, C. A. Insko, and H. S. Ross, "Self-Attribution of Attitude," Journal of Personality and Social Psychology, 17 (March, 1971), 292-297.

L. Ross, M. Lepper, and M. Hubbard, "Perseverance in Self-Perception and Social Perception: Biased Attributional Processes in the Debriefing Paradigm," Journal of Personality and Social Psychology, 32 (November, 1975), 880-892.

Carol A. Scott, "The Effect of Trial and Incentives on Repeat Purchase Behavior," Journal of Marketing Research, 13 (August, 1976), 263-269.

Carol A. Scott, "A Self-Perception Approach to Modifying Socially-Conscious Behavior," Journal of Consumer Research, in press.

Carol A. Scott and Alice M. Tybout, "The Effects of Product Trial, Incentives, and Consensus Information on Consumer Choice," working paper, The Ohio State University, 1977.

Carol A. Scott and Richard F. Yalch, "A Test of the Self-Perception Explanation of the Effects of Rewards on Intrinsic Interest," Journal of Experimental Social Psychology, in press.

C. Seligman, M. Bush, and K. Kirsch, "Relationship Between Compliance in the Foot-in-the-Door Paradigm and Size of First Request," Journal of Personality and Social Psychology, 33 (May 1976), 517-520.

Kelly Shaver, An Introduction to Attribution Processes (Cambridge, Massachusetts: Winthrop, 1975).

Mark Snyder and Michael R. Cunningham, "To Comply or Not Comply: Testing the Self-Perception Explanation of the 'Foot-in-the-Door' Phenomenon," Journal of Personality and Social Psychology, (January, 1975), 64-67.

Michael D. Storms and R. E. Nisbett, "Insomnia and the Attribution Process," Journal of Personality and Social Psychology, 16 (October, 1970), 319-328.

Shelley E. Taylor, "On Inferring One's Attitudes from One's Own Behavior: Some Delimiting Conditions," Journal of Personality and Social Psychology, 31 (January, 1975), 126-131.

Alice M. Tybout, "The Relative Effectiveness of Three Behavioral Influence Strategies as Supplements to Persuasion in a Marketing Context," Journal of Marketing Research, in press.

Alice M. Tybout and Richard F. Yalch, "On Getting Out the Vote: Effects of Labeling on Voter Turnout," in review.

S. Valins, "Cognitive Effects of False Heart-Rate Feedback," Journal of Personality and Social Psychology, 4 (October, 1966), 400-408.

S. Valins, "Persistent Effects of Information About Internal Reactions: Ineffectiveness of Debriefing," in H. London and R. E. Nisbett (eds.), Thought and Feeling: Cognitive Modification of Feeling States (Chicago: Aldine, 1974).

Richard F. Yalch and Carol A. Scott, "Consumer Response to Initial Product Trial: A Bayesian Analysis," in review.

How is self perception formed?

How does self

What is self perception in psychology?

What does self