Guilt Arousing Marketing Communications: an Unexplored Variable

ABSTRACT - The relevance of guilt arousing marketing communications is developed and a rationale for investigation proposed. A brief review of the literature and a discussion of the possible inadequacy of related paradigms follows. A "working draft" conceptual model and several individual difference variables believed to be dominant are then presented. Several research questions in the form of broad hypotheses and possible approaches to measurement are offered to guide preliminary research.

Citation:

Morry Ghingold (1981) ,"Guilt Arousing Marketing Communications: an Unexplored Variable", in NA - Advances in Consumer Research Volume 08, eds. Kent B. Monroe, Ann Abor, MI : Association for Consumer Research, Pages: 442-448.

Advances in Consumer Research Volume 8, 1981      Pages 442-448

GUILT AROUSING MARKETING COMMUNICATIONS: AN UNEXPLORED VARIABLE

Morry Ghingold, Pennsylvania State University

[Special thanks to Dr. R. G. Roden and Peter H. Roden, University of Texas Medical Branch, and all my colleagues at Penn State for their help in the preparation of this paper.]

ABSTRACT -

The relevance of guilt arousing marketing communications is developed and a rationale for investigation proposed. A brief review of the literature and a discussion of the possible inadequacy of related paradigms follows. A "working draft" conceptual model and several individual difference variables believed to be dominant are then presented. Several research questions in the form of broad hypotheses and possible approaches to measurement are offered to guide preliminary research.

INTRODUCTION

Advertising and promotion continue to be central topics within the field of marketing for practitioners and researchers alike. One avenue of inquiry has focused upon the use of emotional appeals in persuasion. Past research dealing with emotional appeals in persuasion has been multi-disciplinary, attracting the interests not only of the scientific community but of special interest and regulatory agencies in the public sector as well.

As has been the case in most other fields, the bulk of research within marketing has centered upon fear arousing communications and the relationship between the level of fear and subsequent shifts in attitude or behavior. Notwithstanding a stream of research which has spanned three decades, the evidence to date has remained equivocal. This may be due in part to disparities in conceptualization or operationalization (Higbee 1969, Ray and Wilkie 1970, Sternthal and Craig 1974).

Surprisingly, however, much less attention has been afforded other plausible emotional appeals. Roseman (1979) has identified 13 distinct emotions believed to span the affective spectrum and many of these (e.g. anger, hope, guilt) seem well suited to persuasive communications. Within a marketing context, a cursory inspection of advertising themes will probably disclose advertising appeals designed to arouse or exploit a variety of emotional responses. One such emotion may be guilt.

The concept of guilt has long been dominant in such diverse fields as theology, philosophy, and counseling and psychiatry, disciplines in which guilt and guilt arousal are viewed as effective agents of attitude and behavior change (see Schneiders 1967, 1968; Stein 1968). However, it would seem that only marginal effort has been expended in examining the role of guilt in persuasive processes, a field of inquiry central to marketers and marketing theorists.

Research scarcity alone, however, is insufficient cause to merit investigation of guilt arousing communications. Preliminary interest could focus on differences existing between guilt arousing communications and other persuasive communications which also use emotional appeals but may differ theoretically, as may be the case with fear appeals.

Another incentive for investigation of guilt arousing communications, albeit less compelling, might be the sheer number of advertisements which appear to have the intention of arousing some degree of guilt in one form or another (for some possible examples of guilt arousing communications in print media see Time 1980a, 1980b, 1980c). Wheatley and Oshikawa (1970) suggest that advertising practitioners tend to prefer positive appeals over negative or anxiety arousing appeals such as fear or guilt. In light of this preference, an impartial observer might cautiously surmise that at least some advertisers believe negative appeals to be effective for their own purposes considering the vast resources allocated for production and dissemination of these messages.

Although perceived (by the author) to be in fairly widespread use in marketing communications, guilt appeals have thus far not been investigated within the discipline. The lack of conceptualization, theory, and research has prevented greater understanding of this enigmatic marketing variable. This paper, therefore, has three purposes: 1) to introduce a new area of interest, 2) to offer a review of the literature dealing with guilt arousing persuasive communications, and 3) to attempt to conceptualize this construct and suggest research on guilt arousing marketing communications.

RESEARCH OVERVIEW

The purpose of this section is to follow the course of research which has dealt with guilt arousing communications, a particular form of negative-emotional/anxiety arousing communications. Following sections will attempt to differentiate between guilt and other negative-emotional appeals, of which fear has been singularly predominant, and to begin development of a "first draft" conceptual model.

A major obstacle which has hampered the development of a clear linkage between guilt appeals and research in the area of negative-emotional or anxiety arousing communications has been the concentration of effort in the domain of fear research. Fear appeals research prospered for more than two decades, and clearly has been the preferred negative emotion or anxiety for investigation. This state of affairs has best been exemplified by Leventhal and Trembly (1968). In their article "Negative Emotions and Persuasion,'' the authors restricted their experimentation and discussion to fear alone, rather than other plausible negative emotions which may have persuasive potential such as guilt, anger or sorrow. Whether these authors believed that a fear paradigm was sufficient to explain the relationship between all negative emotions and persuasion or that all negative emotions were either interchangeably similar with or a subset of fear remains unclear. Perhaps in light of the absence of competing paradigms for other emotions or anxieties their approach was totally appropriate. It should, however, have signaled the need for conceptualization and research in new areas.

Wheatley and Oshikawa (1970) reported on the relationship between anxiety and positive and negative advertising appeals. The authors stated that researchers in the social sciences had compiled evidence sufficient to attest to the power of negative appeals and suggest that advertisers may be overlooking or minimizing an important persuasive tool. Citing past research the authors proposed that "negative communication can be effective in inducing the behavior advocated by the communicator (pg. 85)." Applying a general anxiety reduction drive model, Wheatley and Oshikawa expected that reduction of an aroused anxiety would constitute an effective reinforcement and could lead to conformity with a communicator's recommendation. This general model would seem appropriate for a variety of negative emotions or anxieties but as has often been the case only a fear manipulation constituted the negative appeal.

In contrast to many other streams of research, investigation of guilt arousing communications has not evolved through a tradition of theory building. Rather, the emphasis on fear related approaches to research in the areas of anxiety or negative appeals is also evident in the literature on guilt arousal and persuasion. Of the few investigators who have approached the issue of guilt arousing communications, most have seen fit, rightly or wrongly, to graft their research onto existing fear appeal paradigms (Haefner 1956, Zemach 1966, Yinon et al. 1976). So engaged, these researchers could hope for no more than a general level hybrid theory of guilt arousing communications juxtaposed with fear research.

More to the point, attempts at hybrid guilt theory building involved optimally matching empirical findings to existing fear models. The earliest work, by Haefner (1956), involved a manipulation of guilt in two levels, high and low, perhaps in an effort to replicate the "inverse relationship" in fear research reported by Janis and Feshback (1953, 1954). One may interpret Haefner's approach as viewing guilt as a subset or alternate form of fear. Haefner was exploring negative emotional arousal, operationalized as fear and guilt (both of which were significant), then comparing his findings to the earlier Janis and Feshback (1953, 1954) results.

Subsequent investigation of guilt arousing communications reflected changes in the dominant fear paradigm. Zemach (1966) undertook an experimental investigation of the effects of guilt arousing communications, however, the thrust of her research "piggybacked" on fear research. "In view of the paucity of observations and systematic research on guilt arousal, this study was undertaken to investigate whether principles derived from the literature on fear arousing communications might be fruitfully employed in predicting acceptance of recommendations motivated by guilt arousal (pg. 4)." Thus, Zemach's two fundamental hypotheses (the curvilinear "inverse U" relationship between level of guilt arousal and adherence to recommendations, and the dissipation of the differential effects of difference levels of guilt arousal with the lapse of time) were derived from fear research rather than a priori conceptualization. Although both hypotheses were supported, it remains unclear which, if any, theory was supported, the existing fear model or a unique guilt model.

More recently, Yinon at al. (1976) reported an experimental investigation of the effects of guilt arousing communications on acceptance of behavioral recommendations. Citing the competing fear paradigms of Janis (1967) and Leventhal (1970) as evoking "theoretical curiosity of the social psychologists,'' Yinon at al. sought to test Janis' (1967) contention that the curvilinear "inverse U" model also applied to guilt arousing appeals. Although the curvilinear model was supported, Yinon at al. failed to measure any individual difference variables, as Zemach (1966) had done. As well, Yinon at al. reported operational difficulties in effectively manipulating three levels of guilt, leading to a posteriori collapsing of the design, and casting at least some shadow of doubt on the experimental findings.

Kelman (1979) discussed the relationship between various emotional states and attitude change. Although Kelman (1979) did not specifically address guilt arousing communications, he did propose that guilt, as an emotional response, was among the most likely moral dilemmas to lead to attitude change. Changes in attitude may occur toward the object or issue involved in an action or toward the action itself. Kelman proposed that when justifying an action or making it more acceptable were not available as possible changes in attitude, individuals must confront the discrepancy between their actions and role expectations and "may actually change their attitudes in the direction of more favorable evaluation of the object or issue involved in the action (pg. 74)." Kelman stated that these notions were derived from dissonance theory.

In summary, the paucity of guilt appeal research reported by Zemach (1966) has continued to the present. Clearly a stream of research founded upon three studies conducted over a period of twenty years could not be expected to have nurtured much theory. Rather, that research which has been reported has really been in the domain of fear research. Although one cannot dismiss this approach out of hand, the fact that guilt arousal has not received attention as a unique construct unto itself is disappointing. In light of the differences between guilt and fear to be developed below, this paper seeks to argue in favor of viewing guilt arousing communications as a distinct persuasive mechanism employing an emotional appeal theoretically different than fear. As such, a "first draft" conceptual model specific to guilt arousing communications will be proposed.

TOWARD A CONCEPTUALIZATION OF GUILT AROUSING PERSUASION

Guilt as a Psychological Construct

As a psychological term, English and English (1976) define guilt as "a realization that one has violated ethical or moral or religious principles, together with a regretful feeling of lessened personal worth on that account." Wolman (1973) added that with the lowering of self esteem or feeling of lessened personal worth also comes a need to mike retribution for the transgression. So defined, guilt may be viewed as an a posteriori emotional response which follows a particular action or thought, what Wolman (1973) called the transgression. The need to make retribution may be considered as one possible outcome resulting from a desire to reduce the level of guilt feelings to a more tolerable level, assuming some idiosyncratic threshold had been surpassed.

Unlike most other negative emotions guilt may apply equally to positive and negative outcomes (Roseman 1979). Thus, someone who shoplifts in public but is not apprehended may experience guilt even though the outcome of the act was essentially positive. Roseman maintains that both forms of guilt are phenomenologically, physiologically, and behaviorally alike inasmuch as all "guilty" people have done something "wrong" and can expect to be sanctioned by others.

Fear, in contrast, is more an anticipatory emotional response. Fear is the anxiety brought about by anticipated consequences to particular actions or cognitions, in other words an a priori emotional response. Fear results from the perception of some danger and brings about the goal of avoiding or preventing the threatened outcome. Guilt, on the other hand, arises from the perception that an individual has him/herself committed some injustice and prompts behavior designed to undo or atone for the offense (Roseman 1979). Since guilt is brought about by the actual consequences of behavior and fear is caused by the anticipated consequences of behavior, theoretically the two forms of anxiety should be different. Thus, guilt should be considered as a persuasive construct unto itself.

There is, however, a "fuzzy" area of overlap apparent between these and other forms of anxiety. Situations may arise in which it is unclear which emotion is operative or central when an individual senses "fear of guilt" or "guilt from fear" for example. Whether the response is a priori or a posteriori depends on the interpretation of the situation. "Ring around the collar" may be veered as a fear appeal, causing one to seek to avoid or prevent this unwanted outcome, or alternatively as a guilt appeal, prompting the launderer to undo or atone for his/her error. The interrelatedness of anxieties or negative emotions is the cause of the overlap. Roseman (1979) has identified 5 "dimensions" that give rise to emotions and relate discrete emotions to one another in a structural model of affect. Roseman's model of affect can be used to account for perceived similarities among various emotions and predict pat-terms of to-occurrence and sequential occurrence in complex emotional reactions.

Guilt as Cognitive Dissonance

As noted earlier, F. Kelman (1979) argued that guilt was a powerful agent in attitude change. Kelman referred to guilt as one of a number of moral dilemmas. In the case of guilt, attitude change would be in the form or retrospective Justification for an action in the same manner postulated by dissonance theory (Festinger 1957).

Dissonance theory (Festinger 1957) can easily be adapted to the problem of guilt. Feelings of guilt are, for most individuals, clearly dissonant cognitions creating inconsistency in some cognitive and/or perceptual elements. The two central hypotheses of dissonance theory are 1) the presence of dissonance motivates the individual to reduce it, thereby achieving consonance, and 2) when dissonance is present the individual will avoid situations and behavior which will increase it.

There are some problems, however, in applying dissonance theory to the problem of guilt arousing marketing communications. Firstly, dissonance theory is very broad in nature and has been modified by a number of researchers who have specified particular conditions under which dissonance will or will not occur (Calder 1973). "The real problem has been in tying down the theory. The very looseness which invites creative derivations renders the theory ell but impossible to disprove (Calder 1973, pg. 259)." Thus the broad nature of dissonance theory may render it too general to effectively deal with guilt arousing communications at all levels of analysis. Rather, individual responses to guilt appeals represent but a particular form of cognitive dissonance, in much the same way that other anxiety arousing communications and theories of cognitive consistency are related to the general theory of cognitive dissonance.

A second difficulty rests in measurement and operationalization issues. Given the voluminous research in the area of cognitive dissonance, it would seem that there has not been a generally practiced or accepted way of effectively manipulating and measuring varying levels of "aroused" cognitive dissonance, due in part to the abstractness of the construct and in part to its general nature and vagueness since cognitive dissonance can deal with any cognitive elements. In dealing with a particular form of dissonance, such as guilt arousing communications, it may prove easier to variably manipulate the construct and obtain measures of arousal. Clearly measurements in the domain of an anxiety such as guilt cannot achieve perfection, but they may be easier to fashion because of the narrower and better defined boundaries of the construct of interest. (Several possible methods for measuring aroused guilt will be discussed briefly later in the paper.)

If guilt and cognitive dissonance are but similar constructs, and not in fact one and the same, there may be marked differences between an individual's threshold for cognitive dissonance and his/her threshold for aroused guilt, or in the intensity of the individual's drive to return to the desired mental state. Similarly, one might anticipate different resolution strategies for cognitive dissonance and aroused guilt. Although the essential notion is the same, resolving the undesired mental state (dissonance or guilt), the methods available and selected may not be interchangeable or even compatible.

In general, although guilt arousal is a mental state which behaves similarly to the stipulates of dissonance theory, it is unique at the same time. More to the point, dissonance theory is so broad and malleable that it may be adapted to encompass many diverse phenomena since it is concerned with any and all cognitions. Understanding dissonance theory may be viewed as necessary but not sufficient for understanding guilt. The specificity and particularities involved require that guilt be studied as a separate entity even though clear linkages nay be drawn to dissonance theory.

Guilt as Information Processing Conflict

Investigation of guilt appeals may also view aroused guilt as a form of conflict in information processing. Berlyne (1957) described conflict as the result of coveting and incompatible response tendencies. Thus, in the case of choosing between a response tendency of self gratification versus altruism, (i.e., dining out instead of renewing one's donation to the Red Cross) the conflict which may subsequently occur may be analogous to guilt. Often, the guilt or conflict may not even be known to exist until aroused, perhaps as by a communication in a marketing context.

Bettman (1979) cites Hansen (1972) and Howard and Sheth (1969) in suggesting that conflict be viewed as an equilibrium modal. Individuals would be viewed as having desired optimum levels of conflict, above or below which undesired disequilibrium occur. As with cognitive dissonance, the notion of conflict represents a form of inconsistency which would lead to a response aimed at its resolution. Kelman and Baron (1968) proposed a framework of alternative modes for resolving inconsistencies, as those created by conflicts, as well as hypotheses concerning the conditions under which the various strategies might be used.

Although guilt is clearly a form of conflict arising from competing response tendencies, a conflict resolution model would seen not to be generalizable to all guilt arousal situations. However, there is some similarity between certain forms of guilt and the special cases of avoidance-avoidance conflicts and approach-avoidance conflicts. The overlaps between guilt and information processing conflict occur, however, only with special cases of each and not as a general principle and the two are not interchangeable since approach-approach conflict is not analogous to guilt. As well, the notion of an equilibrium model for conflict seems somewhat counterintuitive when generalized to guilt. As a rule it would probably be safe to say that few individuals seek to increase their feelings of guilt because their current mental state is too free to guilt. Rather, a threshold effect appears more logical.

As was the case with cognitive dissonance theory, there are clear linkages between conflict in information processing and a conceptualization of guilt arousal in persuasion. However, there also appear to be several clear distinctions, such as the equilibrium effect and inapplicability of positive response tendencies and it would be inappropriate to assume that information processing conflict and guilt arousal are one and the same phenomenon. Rather, in specific situations and under particular conditions they are quite similar.

In light of the dubious piggybacking on fear research which has occurred in the past, and the apparent inappropriateness of cognitive dissonance and information processing conflict as fully viable theoretical explanations, there is a need for a unique conceptualization concerned with guilt arousal in a persuasive context. A "first draft" of such a paradigm will now be proposed.

A "First Draft' Conceptualization of Guilt Arousing Communications

Given the definition of guilt as a psychological construct noted earlier, several constructs are implied which must be incorporated into a model of guilt appeals. Firstly, a hypothetical mental state of aroused guilt must be assumed to exist. Implicit in such a construct would be the notion of an idiosyncratic threshold, that level beyond which guilt could no longer be tolerated. Since it is likely that a single communication would elicit varying degrees of aroused guilt in different individuals, the construct mental state of aroused guilt should be mediated by the associated construct susceptibility to guilt arousal, although the abstractness of this construct would make it difficult to distinguish and measure.

Assuming a communication has aroused guilt in an individual beyond the tolerance threshold, but below an intolerance threshold, the next logical construct would refer to the drive mechanism of reducing or resolving the feelings of guilt, an effort to restore moral equilibrium between the individual and his/her environment (Roseman 1979). The drive to reduce undesirable feelings of guilt manifests itself in other more abstract constructs, such as beliefs and/or attitudes in cognitive structure, a desire for additional information (possibly for purposes of bolstering or transcendence), or behavioral intent and ultimate behavior for example. It is likely, however, that another mediating variable will impact upon these abstract constructs, that being the individual's susceptibility to persuasion (which in this instance would be related to susceptibility to guilt arousal).

Given that a marketing communication had been judged to arouse guilt in a message recipient, and the message had been duly attended to, the directionality of the proposed paradigm would be as follows. Dependent upon the individual's susceptibility to guilt arousal, a mental state of aroused guilt would be expected to occur. Should the threshold for aroused guilt be surpassed in the mental state the individual would be expected to enter a drive state intent upon reduction or resolution of the evoked feelings of guilt. This drive might then bring about movement, cognitive or behavioral, to resolve the undesired mental state. A reasonable expectation would be that both the mental state of aroused guilt and subsequent resolution tendencies would be related to the individual's susceptibility to aroused guilt since those more susceptible would likely surpass the hypothesized threshold point sooner than individuals with a similar threshold but less susceptibility to guilt arousal.

Although it may be impossible to compile an exhaustive list of resolution strategies, several dominant ones with direct marketing implications bear consideration. Of interest in a marketing context would be resultant changes in beliefs, attitude or more generally, cognitive structure in the direction advocated by the communication. Such changes would depend on the individual's susceptibility to persuasion in the particular context of the communication. For example, an individual with a large latitude of rejection would likely contrast the communication as being too extreme, possibly demonstrating a polarization of attitude rather than attitude change (Sherif and Hovland 1961, Sherif 1979). Behavioral intent and ultimate behavior are also plausible vehicles for manifestation of resolution strategies, although as above, changes would depend upon susceptibility to persuasion, which in the case of guilt appeals would in turn depend upon susceptibility to guilt arousal. A further resolution strategy of consequence for marketers could be a desire for additional information, perhaps as a prelude to reevaluation and subsequent attitudinal or behavioral change.

Similar to models of cognitive dissonance, the proposed paradigm is essentially a threshold model for the hypothetical construct representing the mental scats of aroused guilt. When the level of guilt surpasses the individual specific threshold, resolution of the guilt is desired. However, should the level of guilt be too high, surpassing the threshold of intolerance, it may well be that the persuasive attempt will be without impact. This may be due to possible denial or distortion of the message content to avoid intolerably high levels of guilt, rather than attempting to resolve them. Once the transgression has been resolved (rather than denied or distorted) as with changes in cognitive structure, behavioral intent or behavior, the individual would cease efforts to reduce the feelings of guilt, returning to a mental state in which the level of aroused guilt was below the threshold.

In sum, by arousing feelings of guilt which refer to a particular context and that surpass a given threshold (but not the threshold of absolute intolerance), an individual would be expected to attempt to alleviate such feelings, possibly through changes in cognitive structure, behavioral intent or behavior. If the feelings of guilt were aroused by a marketing communication which included recommendations, implicit or explicit, leading to a reduction of guilt, the recommendations would more likely be adhered to as mediated by susceptibility to aroused guilt and/or persuasion in terms of attitude, behavioral intent or behavior than if the recommendation were posited without the accompanying guilt arousing communication. (A graphic representation of the "first draft" model can be seen in Figure 1.)

SOME POSSIBLE INDIVIDUAL DIFFERENCE VARIABLES

A number of individual difference variables may be hypothesized to mediate the efficacy of guilt arousing communications in a given population. Although many such variables are generalizable to other forms of persuasion, some are more closely linked to the paradigm proposed herein. Since one may conceptualize a great number of relevant individual difference variables, only those considered of significant impact will be discussed at this early stage of model building.

One trait related to many forms of persuasion is locus of control (Rotter 1966). Rotter differentiated between individuals believing in an internal locus of control or an external locus of control. In broad terms, internals perceive themselves to control their own destinies, whereas externals expect control from outside forces. Rotter argued, with the support of empirical evidence, that externals would be more prone to subtle suggestion and persuasion. With respect to guilt arousing marketing communications, one would expect that an external would more likely adhere to the recommendations advocated in the communication as a means of resolving feelings of aroused guilt. An internal might possibly seek his own solution to the moral dilemma. Thus, locus of control may mediate resolution strategies to aroused guilt.

Another general individual difference variable in persuasibility is self esteem. Leventhal and Perloe (1962) found that for certain message types subjects high in self esteem were influenced more by positive communications than by negative communications. Subjects low in self esteem showed the opposite pattern. With respect to guilt arousing communications, Zemach (1966) reported that self esteem was a powerful explanator in her investigation, although it remains unclear whether self esteem mediated overall susceptibility to persuasion alone or susceptibility to guilt arousal as well, since the two are probably correlated and it would be difficult to separate and independently measure the constructs without confounding one with the other.

Goldstein (1959) hypothesized that the acceptance or nonacceptance of recommendations contained in a persuasive appeal would be related to the subject's characteristic reaction to tension-producing stimuli based on Mainord's (1956) distinction between copers and avoiders. Coping and avoiding behavior parallels the later formulation of Kelman and Baron (1968) for handling inconsistency, copers would be expected to exhibit greater impact of the tension-producing persuasive appeal than would avoiders. Thus, with respect to guilt arousing communications, copers should be more susceptible to high levels of aroused guilt and persuasion in general. Following Goldstein's findings, low tension-producing (i.e., guilt arousing) communications would be more effective for avoiders than for copers.

FIGURE 1

A "FIRST DRAFT" MODEL OF GUILT AROUSING COMMUNICATIONS

A final variable which may be postulated to mediate guilt appeal effects would be a trait of self blame or inherent guilt in an individual. This psychological construct has not been applied to the area of persuasion to date, although it appears relevant when considering guilt arousing communications. Storm et al. (1958) developed the construct as well as a scale for its measurement. It would seem logical that individuals with a tendency for self blame may be more susceptible to communication induced guilt, and as such, more likely to react in accordance with the recommendations advocated in a guilt arousing communication. In contrast, it may be postulated that individuals high in inherent guilt possess higher thresholds for aroused guilt, delimiting the persuasive efficacy of guilt arousing communications. Thus, although it would seem that this construct intervenes in the persuasion process, arguments as to the directionality of the influence may support either suppression or facilitation of the guilt appeal effect.

In short, four factors have been proposed as likely individual difference variables which may mediate the effects of guilt arousing communications; locus of control, self esteem, coping and avoidance behavior, and self blame and inherent guilt. Although other variables may come into play, those noted above are likely to be prominent and are therefore worthy of consideration at this preliminary stage of the theory building process.

A RESEARCH AGENDA: SOME PRELIMINARY HYPOTHESES

Although the graphic representation of the model (Figure 1) appears rather elaborate, a single experimental investigation could be sufficient as a preliminary test of the proposed paradigm. The research would be of an exploratory nature, centering upon a phenomenon of which little is known. Of particular interest in the study to be proposed are the following broad hypotheses:

H1: The mental state of aroused guilt will be greater under manipulations of moderate guilt than under minimal or maximal guilt.

H2: The greater the susceptibility to guilt arousal, the greater the level of aroused guilt in the mental state.

H3: The greater the susceptibility to aroused guilt, the greater the likelihood of a drive to resolve feelings of aroused guilt and subsequent susceptibility to the persuasive appeal.

H4: The greater the level of aroused guilt in the mental state, the greater the drive to resolve the feelings of guilt.

H5: The greater the drive to resolve feelings of guilt, the greater the changes in cognitive structure, behavioral intent, or desire for additional relevant information.

Thus, the hypotheses infer a curvilinear relationship between levels of guilt in the communication and the mental state of guilt. This reflects the double threshold effect described above. Susceptibility to aroused guilt is also hypothesized to affect the mental state of guilt, as well as the drive to resolve the feelings of guilt and subsequent susceptibility to the persuasive appeal. The drive to reduce feelings of guilt is hypothesized to be manifested in one of several ways, through changes in cognitive structure (e.g., changes in beliefs or attitudes in the advocated direction or bolstering of existing attitudes), changes in behavioral intent relevant to the message recommendations, or the desire for more information relevant to the message content, although other means may likely exist.

MEASUREMENT OF AROUSED GUILT

Before closing this discussion it may be useful to briefly outline some approaches to the measurement of aroused guilt since the contention of this paper has been that guilt arousing persuasion is an unexplored and enigmatic variable in the marketing discipline. Several methods will be proposed since empirical tests should not rely on a single measure.

Given that an individual has been presented with a guilt arousing communication, a cognitive response format would appear to be the best way to begin. The first question might ask the respondent to elicit (verbally or in writing) any thoughts that the communication brought to mind. A second question might direct the respondent more towards effective responses, asking that he/she elicit any feelings or emotions that the communication made him/her feel. Conceivably these questions could "create" a great deal of data of which only a small fraction would be of interest. Coding and coder reliability would also be problems and the analyst would have to specify a priori which responses were indicative of guilt feelings and/or to what degree. Nonetheless, cognitive responses immediately after exposure might offer the best measures of aroused guilt.

Another method of measurement could be the use of a scaling technique which has respondents indicate the degree to which they feel or experience a particular emotion. Haefner (1956) used a cluster of nine adjectives as an index of guilt (ashamed, blameworthy, conscience-stricken, contrite, guilty, regretful, remorseful, repentant, and sorry) Judged by clinical psychologists to be relevant to that form of emotional arousal. Haefner found that his treatment groups showed substantial shifts on this cluster although he did not factor analyze them. This is essentially the method used by Zemach (1966, pg. D1, D2) to evaluate the effective responses to the communications used in her study, although she did not use all the adjectives in Haefner's cluster. Although open to demand characteristics, this method could be expected to realize useful measures of aroused guilt.

Similar to the above method, semantic differential scales could also be used to allow the respondent a means of expressing how the communication makes him/her feel. Care should be taken to avoid redundancy with any adjectives used in other scales or measures.

Lastly, attitudinal measures could be used to gauge the arousal of guilt relevant to the communication. If an advertisement was used which was intended to arouse guilt feelings over the plight of the world's underprivileged and garner charitable donations, respondents could be asked the extent to which they feel partly to blame for the plight of the poor, or the extent to which they regret not having done more to help the underprivileged. These measures are more specific to the message content but should also yield some indication of the level of guilt aroused by the communication.

In sum, several approaches to measuring aroused guilt may be used; cognitive responses, affect scales, semantic differentials, and attitudinal measures. It would be advisable to incorporate multiple measures in an experimental design to offer at least some opportunity for tests of validity.

IN CONCLUSION

This paper has proposed that guilt arousing communications are frequently occurring phenomena in a marketing context and worthy of investigation by consumer researchers. The insufficiency of past explorations into guilt appeals and the presumed differences between guilt and related paradigms of fear, cognitive dissonance and information processing conflict provided a rationale for more focused research. Constructs were explicated and an initial "first draft" model was presented. Several individual difference variables were proposed and a research agenda was outlined in the form of broad preliminary hypotheses and approaches to measurement issues. No doubt this paper has raised more questions than it has answered but it is hoped that this first attempt at conceptualization of the phenomenon will arouse interest and stimulate additional efforts to add increased understanding and control of this exciting marketing variable.

REFERENCES

Berlyne, D. E. (1957), "Uncertainty and Conflict: A Point of Contact Between Information-Theory and Behavior-Theory Concepts," Psychological Review, 64 (November), 329-339.

Bettman, J. R. (1979), An Information Processing Theory of Consumer Choice: Addison Wesley Publishing Company.

Calder, B. J. (1973), "Cognitive Consistency and Consumer Behavior," in H. U. Kassarjian and T. S. Robertson, (eds.), Perspectives in Consumer Behavior, Glenview, IL: Scott, Foresman and Co., 2&7-263.

English, H. B. and English, A. C. (1976), A Comprehensive Dictionary of Psychological and Psychoanalytical Terms, New York: David McKay Co.

Festinger, L. (1957), A Theory of Cognitive Dissonance, Stanford: Stanford University Press.

Fishbein, M. and Ajzen, I. (1975), Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research, Reading: Addison-Wesley Publishing Company.

Goldstein, M. J. (1959), "The Relationship Between Coping and Avoiding Behavior and Response to Fear-Arousing Propaganda," Journal of Abnormal and Social Psychology, 58 (March), 247-252.

Haefner, D. P. (1956), Some Effects of Guilt Arousing and Fear Arousing Persuasive Communications on Opinion Change. unpublished technical report.

Hansen, F. (1972), Consumer Choice Behavior: A Cognitive Theory. New York: Free Press.

Higbee, K. (1969), "Fifteen Years of Fear Arousal: Research on Threat Appeals, 1953-1968," Psychological Bulletin, 72 (December), 426-444.

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

Janis, I. L. (1967), "Effects of Fear Arousal on Attitude Change: Recent Developments in Theory and Experimental Research," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, New York: Academic Press, 167-225.

Janis, I. L. and Feshback, S. (1953), "Effects of Fear-Arousing Communications," Journal of Abnormal and Social Psychology, 48 (January), 78-92.

Janis, I. L. and (1954), "Personality Differences Associated with Responsiveness to Fear-Arousing Communications,'' Journal of Personality, 23 (December), 154-166.

Kelman, H. C. (1979), "The Role of Action in Attitude Change," paper prepared for the 1978-79 Nebraska Symposium on Motivation, Department of Psychology, University of Nebraska, Lincoln,

Kelman, H. C. and Baron, R. H. (1968), "Determinants of Modes of Resolving Inconsistency Dilemmas: A Functional Analysis," in R. P. Abelson et al. (eds.), Theories of Cognitive Consistency: A Sourcebook, Chicago: Rand McNally, 670-683.

Leventhal, H. (1970), "Findings and Theory in the Study of Fear Communications,' in L. Berkowitz (ed.), Advances in Social Psychology, New York: Academic Press, 119-186.

Leventhal, H. and Perloe, S. (1962), "A Relationship Between Self Esteem and Persuasibility," Journal of Abnormal and Social Psychology, 64 (June), 385-388.

Leventhal, H. and Trembly, G. (1968), "Negative Emotions and Persuasion," Journal of Personality, 36 (March), 154-168.

Mainord, W. A. (1956), Experimental Repression Related to Coping and Avoidance Behavior in the Recall and Relearning of Nonsense Syllable, unpublished doctoral dissertation, University of Washington.

Ray, M. L. and Wilkie, W. L. (1970), "Fear: The Potential of an Appeal Neglected by Marketing," Journal of Marketing, 34 (January), 54-62.

Roseman, I. (1979), "Cognitive Aspects of Emotion and Emotional Behavior," paper presented at the 87th Annual Convention of the American Psychological Association, New York City.

Rotter, J. B. (1966), "Generalized Expectancies for Internal Versus External Control of Reinforcement," Psychological Monographs (Whole no. 609), 1-28.

Schneiders, A. A. (1967), "Catholics and Psychology," Catholic Psychological Record, 5 (2), 102-110.

Schneiders, A. A. (1968), "The Nature and Origins of Guilt," Transactions of the New York Academy of Sciences, 30 (5), 705-713.

Sherif, C. W. (1979), "Ego Involvement and Attitude Change," paper prepared for the 1978-79 Nebraska Symposium on Motivation, Department of Psychology, University of Nebraska, Lincoln.

Sherif, M. and Hovland, C. I. (1961), Social Judgment Assimilation and Contrast Effects in Communication and Attitude Change, New Haven: Yale University Press.

Stein, E. V. (1968), Guilt: Theory and Therapy, Philadelphia, PA: The Westminster Press.

Sternthal, B. and Craig, C. S. (1974), "Fear Appeals: Revisited and Revised," Journal of Consumer Research, 1 (December), 22-34.

Storm, T., Rosenwald, G. C. and Child, I. L. (1958), "A Factor Analysis of Self-Ratings on Social Behavior," Journal of Social Psychology, 48, 45-49.

Time (1980a), advertisement appearing on page 9, Vol. 115 (5).

Time (1980b), advertisements appearing on pages C2 and 58, Vol. 115 (7).

Time (1980c), advertisement appearing on page 17, Vol. 116 (1).

Wheatley, J. J. and Oshikawa, S. (1970), "The Relationship Between Anxiety and Positive and Negative Advertising Appeals," Journal of Marketing Research, 7 (February), 85-89.

Wolman, B. B. (1973), Dictionary of Behavioral Science, New York: Van Nostraud Reinhold Company.

Yinon, Y., Bizman, A., Cohen, S. and Seger, A. (1976), "Effects of Guilt-Arousal Communications on Volunteering to the Civil Guard: A Field Experiment," Bulletin of the Psychonomic Society, 7 (6), 493-494.

Zemach, M. (1966), The Effects of Guilt-Arousing Communications on Acceptance of Recommendations, unpublished doctoral dissertation, Yale University.

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