The Roles of Emotion in Consumer Research

Elizabeth C. Hirschman, Rutgers University Barbara B. Stern, Rutgers University ABSTRACT

The aim of this paper is to describe and synthesize findings across three different setsofliterature dealing with emotion: (1) the cognitive stream, (2) the hedonic consumption stream, and (3) the compulsive/addictive consumption stream. We present a model of emotion and suggest novel directions for future inquiry.

INTRODUCTION
A decade and a half have passed since consumer researchers began attending to the roles of emotion in consumption (Hirschman and Holbrook 1982, Holbrook and Hirschman 1982); and over a decade has elapsed since Gardner's (1985) integrative review of mood states. During this time period an extraordinary level of research has occuned within the boundaries of our discipline regarding emotion. The aim of this paper is to describe and synthesize these findings and to present a inclusive model of emotion and its influence on consumer behavior. In the course of thiseffort we will review three sets of literature dealing with emotion. These include (1) the cognitive stream, typified by inquiries regarding emotion's role as the affective component of attitude (e.g., Aaker and Stayman 1989) or as a precursor to cognitive restructuring (Isen 1984); (2) the hedonic consumptionstream asexemplified early on by the work of Holbrook and Havlena (Havlena and Holbrook 1986) and most recently by Celsi, Rose and Leigh (1993) and Amould and Price (1993); (3) the computsiveladdictive consumption stream beginning with early work on impulse purchasing by Rook (198S) and evolving into more general models as exemplified by Faber and O'Guinn (1989), Hirschman (1992) and Rindfleisch, Burroughs and Denton (1997).

also feed-back into emotion; for instance, thinking about the death of a pet may cause the individual to become sad. Finally, the model proposes that consumption behaviors occurring in response to consumers' emotional states are shaped by prevailing cultural norms. For example, a white, middle class woman may 'treat' her chronic unhappiness with compulsive shopping and bulimia, while a white, middle class man may deal with his depression by drinking alcohol, and a white, middle class teenager may respond to his/her 'blues' by smoking marijuana. We begin a more detailed discussion of the model with the research stream most closely aligned with emotion in consumer research: cognitive psychology. Cognitive Psychology and Emotion An examination of this literature base rightly begins with Meryl Gardner's comprehensive (1985) treatise, 'Mood States and Consumer Behavior: A Critical Review.' Gardner's review focused upon moods, defined as 'feeling states that are transient;... particularized to specific times and situations(1985, p. 282).' Thus, her review did not examine research on more intense levels of emotion or on personality dispositions, such as optimism/pessimism, that are relatively stable and enduring characteristics of the consumer. Indeed, as Gardner notes, exogenous mood effects are markedly short lived: 'The effect of a positive mood induction on the performance of a helping task has been found to last approximately twenty minutes' (1985, p. 289). The findings summarized in her review would be relevant to the portion of our model titled 'Current Personal Events' and include those minor events, such as exposure to a pleasing or offensive advertisement, which create a short-term shift in the consumer's prevailing mood. Mood and Advertising EffectivenessS'incc the publication of Gardner's review, six articles have appeared in the Journal of Consumer Research which examined the role of such short term moods in influencing advertising effectiveness. The first, by Holbrook and Batra (1987) sought to incorporate a wider range of emotional intensity in understanding consumers' responses to advertisements and assumed homogenous responses on the part of consumers toward given, emotion-stimulating advertisements. The authors' findings supported the earlier model of emotional content suggested by Mehrabian and Russell (1974) which categorized emotions according to their positive or negative valence and level of activation or arousal. A second study, by Edell and Burke (1987), demonstrated that '/ee//igsgenerated by an ad are differentconceptually from thoughts about the ad and both are important and contribute uniquely to explaining the effects of advertising' (1987, p. 421). The significance of this study was that it demonstrated the independence of mood and cognition in influencing advertising response, as advocated by 2^jonc (1980); whereas some cognitive approaches had argued that mood was merely a subset of cognitive response. Edell and Burke also found that individual consumers varied greatly in the types of emotional responses they had to particular advertisements, noting (p. 431) 'one person's warm commercial may be another person's cool commercial.' This finding was contrary to Holbrook and Batra's (1987) assumption that consumers would exhibit homogeneous emotional responses to given advertisements.

A NEW MODEL
As shown in Figure One, we propose that every consumer has a personal emotional range which can move from extreme happiness to extreme sadness. While most consumers usually experience motions in the middle range, a significant proportion are likely to experience emotions which lie beyond the middle range (Larsen and Emmons, 1985). Indeed, if we view emotion as a normally distributed human variable, then approximately 30 to 50% of the consuming population probably experience emotions above or below the middle range on a regular basis (Larsen and Diener, 1987 a,b). The model further proposes that current personal events of both a major and minor nature will have transient effects on adult consumers' emotional states. A major negative event, such as job loss or divorce, or a major positive event, such as a job promotion or marriage, may cause long-term (say six months) emotional shifts beyond the middle range, while minor positive or negative events (say a helpful salesperson or an offensive advertisement), may only affect emotional states for a few minutes. The model also proposes the direct influence of emotion on a variety of cognitive responses (such as attitude formation and recall) and, thence, on a broad array oiconsumption behaviors, such as impulse purchasing, consumption creativity, and innovativeness. It suggests that consumption behaviors may also feed-back into cognition and emotion; for example, a consumer may discover that a new restaurant's food is delicious, causing a positive attitude to be formed and a happy feeling to occur. And cognitive responses may

Advances in Consumer Research Volume 26, 1999

Advances in Consumer Research (Volume 26) / 5

HGUREl
A Model of Emotional Response Cognitive Responses e.g. attitude formation and change, cognitive complexity

Personal Emotional Range

Current Personal Events

Consumption Behaviors e.g. impulse purchasing, compulsive buying, novelty seeking, gift giving

I
I

Prevailing Cultural Norms

e g. age, gender, race

In another 1987 article, Goldberg and Gom showed consumers a funny or sad television program to induce a positive or negative mood. They then showed them ' informational' versus 'emotionar commercials. Consistent with several other studies in the social psychology literature, they found that both types of commercials were rated as more effective when embedded within the funny program and that this effect was strongest for the emotional commercials. They also found that recall of material contained in the commercials was enhanced under the funny program viewing condition. These findings are consistent with an earlier study by Isen (1984), which suggested that a positive mood enhances information processing ability and recall. Two closely related studies examined the effect of mood on attitudes toward the advertisement (Stayman and Aaker 1988; Batra and Stayman 1990). The first looked at three types of mood warmth, amusement, and irritationand found that mood had a direct effect on brand attitude at low exposure levels, but that this effect appeared to be mediated by A^^ at higher exposure levels. The second study reported that positive moods indirectly influenced consumers' attitudes toward the brand by, first, reducing the number of negative thoughts generated and, second, by reducing total cognitive elaboration, making processing 'more heuristic than systematic' (p. 213).

A 1991 study by Olney, Holbrook and Batra examined how mood responses to an advertisement affected viewing time of the advertisement. They found that advertising content influenced both the positive-negative and arousal dimensions of mood and, not surprisingly, that consumers spent more time watching advertisements with positive content, which created happy moods in the viewer. To sum up, studies of the influence of mood on advertising effectiveness support the general categorization of emotions as positive/negative in valence and high/low in arousal. Multiple studies have found that a positive mood enhances the rating of an advertisement's effectiveness, consumers' information processing ability and recall. Perhaps the most controversial findings relate to the role of positive mood in enhancing information processing ability. Although Isen found that this ability is enhanced by positive mood, Batra and Stayman found that cognitive elaboration is reduced and processing is more heuristic in positive mood states. Thus, the relationship between mood and information processing skill is still not fully understood. Mood Effects in Non Advertising Contexts-Tvfo studies have examined mild emotional responses, i.e., mood, to stimuli aside from advertisements. Swinyard (1993) investigated pre-existing mood states' effects upon consumers' shopping attitudes and inten-

61 The Roles ofEmotion in Consumer Research

tions. Using hypothetical scenarios, Swinyard found that a positive/ negative mood did not have any effect on reported shopping intentions directly, but did act through involvement level (high/ low) to influence shopping intentions. Using mood also as a dependent variable, Swinyard found that a positive/negative shopping experience had a significant effect on mood. This was especially true for highly involved shoppers. As Swinyard (1993, p. 278) reports: 'For involved shoppers a good shopping experience is even better, and a bad shopping experience is still worse.' A second study, Murry and Dacin (1996), investigated how various moods created by television programs affect consumers' liking of the programs. As these authors note, 'Existing consumer theories offer little insight into why consumers seek out... entertainment products such as music that evokes sadness, books that elicit fear, or movies that create anger (1996, p. 439).' The authors found that a negative mood stimulated by a program had a negative effect on program liking, when the program was seen as personally relevant (e.g., 'could happen to me') and realistic. However, programs causing negative moods could still be liked, as long as viewers saw them as irrelevant and unrealistic. Underlying these findings is the larger proposition that consumers will seek out scary or sad consumption experiences only so long as these are not seen as personally threatening. This last study takes us to a different segment of the model shown in Figure One. Under Consumption Behaviors we have listed thrill seeking/risk aversion as a construct relevant to consumption choices. Murry and Dacin's (1996) work suggests that consumers may seek out television programs that provide thrills, so long as genuine emotional risk is avoided. As we shall discuss subsequently, consumers' willingness to take emotional risks is dependent upon what part ofthe emotional range they areJunctioning within. We will propose that consumers who have positive emotional states they believe to be durable will be willing to take more emotional risks. Conversely, those consumers who believe themselves to already be in substantial emotional jeopardy, i.e., are teetering on the edge of depression, will make choices aimed at reducing emotional risks. Isen's Research on Emotion-'We tum now to a subset of the cognitive literature on emotion which hasjustly had a large impact on consumer research investigation of this topicthe work of Al ice Isen. In 1985, Isen, Johnson, Mertz and Robinson examined whether positive affect could influence the contents and complexity of working memory. At the outset to their study, Isen et al. reviewed an extensive literature which suggests that when persons are in good moods, they are typically able to bring more extensive and complex material into working memory, to discern unusual relations between elements in stimulus materials, to have enhanced recall of distantly related items, and to be able to generate a larger number of unusual word associations. As Isen et al. point out, all of these behaviors are associated with creativity, and, indeed, some are even analogous to accepted measuresof creativity (see e.g., Mednick and Mednick 1964). In keeping with this series of prior findings, the Isen et al. (198S) study found that 'the results of this study confirm that unusual [word] associations are to be found from persons who are feeling happy (p. 1418).' These findings are supportive of the proposed model. Several propositions Isen (1987) puts forward in a review of the literature are also consistent with the model we are proposing. First, she notes that a 'large body of research indicates that positive affect can influence social behaviorin particular, sociability, cooperativeness in negotiation, and kindness (p. 206).' Second, happy persons are more likely to be self-indulgent and to give themselves 'self gifts' (Isen p. 210). Thus from Isen's work we find support for various portions of our proposed model. For example, we can support the notion that emotion influences cognitive responses, such as recall, working memory capacity and working memory complexity. We can also support the proposition that emotion, working through cognition, can influence behavior, for example, sociability, gift giving, and helping others. Two additional studies by Isen in creative problem-solving were conducted with medical students and physicians. The first (Isen, Rosenzweig, Young 1991), found that medical students in a positive mood produced accurate patient diagnoses more rapidly than those in a neutral condition and also went beyond the assigned task to consider the medical materials more integratively, working with larger 'chunks' of related information. In the second study (Estrada, Isen, Young 1995), physicians conducting diagnoses under positive affect conditions exhibited this same pattern and, additionally, were found to be less anchored to preliminary assumptions and more open to new information as a reasoning style, than controls. Thus, happy emotion stimulates rapid, accurate cognitive processing, increases the volume of relevant material being processed and enhances cognitive receptivity, all very valuable characteristics for consumers to possess. We conclude this section on cognitive theory with an article by Kahn and Isen (1993) which segues nicely into our next set of literature. In this study Kahn and Isen proposed that (p. 257): 'The induction of |X}sitive affect in consumers carrying out brand-choice tasks may lead to their showing a greater preference for exploration and trying new things in safe and enjoyable contexts. This is because... positive affect improves a person's expectations about the likely outcome of anticipated neutral or positive experiences or events and also prompts people to engage in more elaboration and thinking about neutral things in which they are interested.' And, over a series of three experiments, they found these expectations confirmed. Thus, a happy mood increases consumers' efforts at variety seeking among safe, pleasant alternatives. We next consider the larger literature of which much variety seeking behavior is a partthat of hedonic/experiential consumption.

HEDONIC CONSUMPTION
In 1982 two papers appeared which encouraged consumer researchers to explore the experiential boundaries of consumption beyond the confines of cognitive psychology and utilitarian economics (Hirschman and Holbrook 1982; Holbrook and Hirschman 1982). Following this, several studies were conducted exploring various aspects of hedonic and experiential consumption. We will review five representative examples here. A1986 article by Havlena and Holbrook compared two classification systems for emotion: Mehrabian and Russell's (1974) paradigm and Plutchik's eight emotional categories (1980). Usingan elaborate series of multivariate procedures, they determined that the three dimensional model pleasure, arousal, dominanceproposed by Mehrabian and Russell was superior in representing emotion during actual consumption experiences. A second study, by Allen, Machleit and Klein (1992) examined whether recollected emotions regarding an emotionally-mixed consumption experience (blood donation) could serve as incremental predictors (beyond attitudes) of behavior. They argue that emotion extends beyond attitude and encompasses 'a richer and more diverse domain of phenomenological experience (p. 494).' They further propose that much emotional experience is likely stored in epuex/icmemory, whereas attitudinal judgmentsarelikely found in semantic memory. Using Izard's (1977) emotional tax-

Advances in Consumer Research (Volume 26) 17 onomy the researchers found that 'emotion can have a direct 'compulsive consumption.' The particular type of compulsive influence on behavior that is not captured or summed up by attitude consumption Rook was studying, impulse purchasing, has imporjudgments (p. 500).' This, ofcourse, has implications for our model tant emotional and behavioral com|K)nents. Impulse purchases are in Figure One and suggests that we should draw a causal arrow from associated with happy emotional states, when the consumer is Emotion to Consumption Behavior directly, as well as retaining the feeling self-indulgent, optimistic, enthusiastic, and venturesome. one which already passes from Emotion to Cognition to Behavior. Indeed, much of the safe, pleasant variety seeking behavior as Toward the end of their article, Allen et al. (1992), write described by Kahn and Isen (1993) as well as novelty seeking evocatively of the measurement problems inherent in using emo- behavior (Raju 1980) is likely to occur in an impulse purchasing tion measurement scales and recollected events (p. 502): 'While contest (see Rook 1987, p. 194,196). defending our selection of measures, we do not mean to suggest that A distinction can be drawn between impulse purchasing and "conventional" measures of emotion cannot be improved on for compulsive purchasing, which O'Guinn and Faber (1989) describe consumer research. If we are to do a better job of integrating as repetitive purchasing which is 'ultimately destructive financially emotive experience into consumer research, there are important and psychologically to the consumer, but which provides momenimprovements needed in the measurement area". tary relief from tension, anxiety or sadness (p. 148)'. Impulse The next year, 1993, two articles appeared which went a great purchases usually occur during 'high' emotional states, while distance toward responding to these issues. What was needed was compulsive purchases usually occur as attempts to alleviate 'down' not only novel approaches to measurement, per se, but rather novel emotional states. And, as O'Guinn and Faber (1989, p. 148) have approaches in the methodology used to comprehend emotion within described, compulsive purchasing is closely related biochemically, the consumption experience. We next examine two revolutionary psychologically and behaviorally, to a host of other compulsive efforts which did just this: Amould and Price (1993) and Celsi, consumer activities: "These include substance abuse and extreme excesses in behavior, such as eating disorders and compulsive Rose and Leigh (1993). The context of Amould and Price's research was Whitewater sexuality, as well as more marketplace-oriented behaviors, such as rafting on the Colorado Rivera consumption experience that compulsive gambling and kleptomania". usually evokes the gamut of emotions for most participants. In In their study of compulsive buyers, O'Guinn and Faber contrast to prior studies we have reviewed, the Amould and Price combined depth interviews with survey data to construct a comprearticle used a combination of ethnography and survey-scaling hensive portrait of the phenomenon. They found that most compulmethodologies, creating a very rich and detailed data base from sive shoppers were female (supporting the portion of our model which to construct interpretations. As they comment, the consum- which proposes that cultural norms about gender help direct emoers of the river rafting experience typically have emotional re- tion-laden consumer behaviors, i.e., women are socialized to shop), sponses which are long-lasting and profound: Because ofthe nature and that compulsive buyers tended toward obsessive-compulsive of their methodology, Amould and Price were able to trace socio- traits, had lower self-esteem (both a cause and consequence of logical/anthropological aspects of the emotional experience which compulsive buying), had higher imagination levels and higher traditional measurement devices would have missed. Among these levels of envy and non-generosity (i.e., hoarding) than other conwere the perception by several consumers of the experience as a sumers. And, of course, their behavior had a strong, emotional pilgrimage/rite of passage and the growth of bonding among componentelatedly high when it was occurring, depressingly participants and between the participants and their guide over the low once one came to one's senses. course of the joumey. The research project also shed light on rarely In two later studies, (Faber and O'Guinn 1992; Faber, investigated emotions such as transcendence and ecstasy, which lie Christenson, De Zwaan and Mitchell 1995), these researchers and at the extremes of the positive emotional spectrum. Notably, many theirassociates developed aclinicalscreenerfor compulsive buyers consumers did not want to translate their experience into cognitive and investigated the comorbidity (i.e., co-occurrence) of compulmaterial. sive buying and binge eating. They found that both forms of Similar findings resulted from the investigation of skydiving compulsive consumption occur along a continuum from normal to by Celsi, Rose, and Leigh (1993). The authors proposed that high moderate to severe, a point made by Rook (1987) and Hirschman risk consumer behaviors are 'motivated by a dramatic woridview (1992), as well. In discussing their findings, they reported that (p. 2 ) ' , and discovered that participants' motives varied from thrill gender segregation was again apparent; that is, women tended seeking to social compliance (e.g., to accompany a friend or spouse) toward binge eating, kleptomania and compulsive buying, while and that all participants saw it as involving some degree of risk men predominated in pathological gambling and hard drug usage, acceptance. As with Amould and Price's consumers, these also further supporting the 'Cultural Norms' segment of our model. cited personal mastery (i.e., 'I can do it'), transcendence/ecstasy They also reported results from other studies suggesting that and identity construction/renewal as significant outcomes of the compulsive buyers had greater mood swings in response to buying, behavior. And, as before, the participants felt a sense oicommunitas than normal consumers: 'higher' during purchase, 'lower' after purchase. or bonding with other participants. These latter researchers introduced another aspect of conIn 1992, Hirschman used depth interviews with present and sumer desire for such high risk experiencesthe addictive high recovering drug addicts/alcoholics to construct a general model of resulting from the adrenaline rush they provide. As they note, while compulsive consumption. Her research also included introspective the addiction model does not fully account for all aspects of the accounts of the phenomenology of compulsive consumption. high-risk consumption experience, it does serve as a phenomenoHirschman (1992), like O'Guinn and Faber (1989, 1992, logical explanation for much of it. It also serves as a bridging 1996), noted the serial and/or multiple outlet aspect of compulsive concept to take us to our next set of literature: consumer research consumption. It is posited that because compulsive consumption is conducted on compulsive consumption. rooted in abnormal emotional reactivity, (see Larsen and Diener 1987 a,b) it is an ongoing feature of these consumer's lives. COMPULSIVE CONSUMPTION Typically, they will use multipleforms of compulsive consumption Rook's 1987article,'The Buyinglmpulse'wasoneof the first or tum to a series of compulsive consumption activities over time published in consumer research to investigate what is now termed to allieve anxiety, tension and/or depression. The type and pattem

8 / The Roles of Emotion in Consumer Research FIGURE 2 Emotional Quadrants Positive valence

A
contentment tranquility serenity placidity

B
exuberance delight ectasy elation High arousal

Low
arousal

C
hopelessness dread melancholy lethargy

D
anger hostility panic paranoia

I
of activities these consumers engage in is often guided by social norms regarding what is appropriate for their gender, age, social class and ethnicity. And, as Rook and Fisher (1995) recently found, some types of impulsive consumer behavior may be held in check by social norms, as well. In 1996, DeSarbo and Edwards published a large-scale, survey-based research study that distinguished between impulse purchases and compulsive purchasing. These researchers demonstrated that compulsive buyers react to stress by experiencing higher than average levels of anxiety, i.e., are emotionally reactive. To escape these anxieties, they may tum to compulsive buying, which temporarily provides positive emotions; in essence, compulsive consumers axe self-medicating their anxiety with the emotional high of shopping. DeSarbo and Edwards (1996) additionally found that persons coming from abusive-and/or addictive families were more likely to become compulsive consumers, consistent with O'Guinn & Faber's (1989) and Hirschman's (1992) findings, and supporting the Parents and Childhood/Juvenile Environment segments of our model. Analogously, Rindfieisch, Burroughs and Denton (1996) found that young adults whose parents were divorced or separated exhibited higher levels of compulsive buying.

Negative valence

emotional responses will be filtered through various cognitive activities (e.g., attitude change) and result in a wide-ranging set of consumer behavior, such as impulsive product purchases, variety seeking, and compulsive buying. Which specific behaviors are undertaken will be influenced by prevailing cultural norms regarding what is appropriate for one's gender, age, race, and socioeconomic status. We can also extend this model into a predictive set of propositions and a classification scheme as shown in Figure Two, which is adapted from Holbrook and Batra (1987). The Figure depicts four quadrants resulting from the orthogonal axes of positive-negative emotional valence and low-high emotional arousal.

CONTENTED CONSUMERS/CALM CONSUMPTION

Quadrant A contains the emotions typical of positive affect and low arousal: contentment, tranquility, serenity and placidity. Given what we have leamed from prior research, this section of the emotional space may be labeled as containing Contented Consumers/Calm Consumption. It is likely that some consumers spend much of their lives in this range of emotional experience, and that all consumers experience these emotions from time to time (see e.g., Holbrook and Batra, 1987; Larsen and Diener 1987a). The following speculative propositions^ may be put forward regarding consumers/consumption in this portion of the emotional range.

WHERE DO WE GO FROM HERE?

Our model proposes that adult consumers will exhibit a Personal Emotional Range. Our model additionally suggests that adult consumers will respond emotionally to both major and minor current personal events, such as job loss, exposure to advertising and salesperson encounters. The model further suggests that these

lWe are skating on thin ice here, empirically. Yet we optimistically anticipate these propositions will be confirmed in the near future.

Advances in Consumer Research (Volume 26) 19 1) Consumers experiencing these emotions are contented with the present state of affairs. They are likely to be brand loyal, due to feelings of product/service satisfaction. Because of their low arousal levels, they do not have the energy or desire to be venturesome, innovative or varietyseeking, and they are unlikely to voice complaints or express dissatisfaction. 2) Contented consumers are likely to be quite receptive to 'warm' advertisements (e.g., Aaker and Stayman), as these would help maintain and reinforce their emotional tranquillity. 3) Because Contented Consumers are in a state of low arousal and positive emotion, they may use heuristic processing to simplify decision tasks (see Batra and SUyman 1990), thus avoiding expending energy on cognitive activity. Following from #3 above, it is likely that Contented Consumers will choose to avoid extended problem solving activities; in general, we expect that they will try to avoid activities that are cognitively or physically demanding, because these would disrupt their sense of serenity. These propositions are consistent with some mood experiments, which have found subjects unwilling to undertake activities that might threaten their positive affect, see e.g. Gardner 1985. Happy Consumers/Active Consumption As we move into Quadrant B, we reach feelings such as exuberance, delight, ecstasy and elation. These are the high arousal, positive emotions. Because, as Isen's extensive work has shown, the mental conditions responsible for creating elation and joy are also linked to the stimulation of cognitive and physical activity, many consumers who experience strong positive emotions will feel physically and mentally energized, as well (see also Larsen and Diener 1987b for psychological empirical confirmation). We anticipate that such consumers will be optimistic about consumption endeavors and enthusiastic about acquiring pleasurable experiences. Thus, we propose that: 1. Happy consumers will actively seek out new consumption information, experiences, and products, so long as these are deemed likely to provide pleasure. Following from Kahn and Isen (1993), we expect happy consumers to be innovative, novelty-seeking, and variety-seeking. 2. Consumers who are experiencing delight and elation will have greatly enhanced cognitive abilities. They will be able to construct elaborate cognitive patterns, have increased recall and accelerated processing speed, consistent with Isen's extensive research. 3. As a result of 1 and 2 above, elated consumers should display high levels of consumption creativity: i.e., the ability to use products in novel and unusual ways in order to solve problems. Sad Consumers/Passive Consumption Moving into Quadrant C, we encounter consumers with low levels of arousal coupled with negative emotion: this space is aligned with emotions such as hopelessness, dread, melancholy and lethargy; these persons may be termed Sad Consumers who are likely to engage in Passive Consumption. This area of emotional response has been very little investigated in consumer research. Nonetheless, from the literature on depression and dysthymia (see Isen 198S), we propose that: Consumers who are experiencing the emotions of hopelessness and dread will be very risk averse and pessimistic about consumption activities. They will be unlikely to try new products or services and are at a low probability for variety seeking. 1) Givenl,above,suchconsumersarelikelytodisplaybrand loyalty and habitual purchasing patterns, because they lack the cognitive ability to process novel information or to combine existing information in novel ways. 2) Decision-making by such consumers will be slow, hesitant and aimed toward minimizing possible disappointment. These consumers will stick with 'tried and true' alternatives. 3) Because they are physically and mentally lethargic, sad consumers will be less likely to express dissatisfaction or voice complaints, even when product performance is unsatisfactory. They are likely to believe that such activities are 'not worth the effort' and would be futile. 4) Sad consumers' behaviors will be marked by simplicity, ritualism, passivity and resignation. They are based on the premise that 'things will never get any better than this.' Decision-making by such consumers will be slow, hesitant and aimed toward minimizing possible disappointment. These consumers will stick with 'tried and true' alternatives. Because they are physically and mentally lethargic, sad consumers will be less likely to express dissatisfaction or voice complaints, even when product performance is unsatisfactory. They are likely to believe that such activities are' not worth the effort' and would be futile. Sad consumers' behaviors will be marked by simplicity, ritualism, passivity and resignation. They are based on the premise that 'things will never get any better than this.' Angry Consumers/Hostile Consumption As we continue toward Quadrant D, we cross-over into the high arousal/negative emotions anger, hostility, panic and paranoia are the feelings often experienced by persons here. It is likely that anger and hostility are the most commonly experienced consumption emotions within this portion of the spectrum; for example, a delayed flight, a malfunctioning car, or a botched repair job can sometimes provoke rage in consumers. Again, little empirical consumer research has been conducted on these emotions, but some speculative propositions can be made: 1. Obviously, consumers with tendencies toward active dysthymia are those most likely to 'lose their temper' in a provocative consumption situation, but all consumers occasionally find themselves in this emotional space. Loud, demanding, complaining behavior is I ikely to occur, perhaps even including physical damage to property or personnel (e.g., assaulting a waiter). Research on how to disarm such volatile negative consumer behaviors would be very valuable. 2. Drawing from the compulsive consumption literatures (e.g., O'Guinn and Faber 1989), we propose that much

10 / The Roles of Emotion in Consumer Research drug and alcohol abuse, compulsive eating, compulsive shopping, pathological gambling and similar addictive/ compulsive behaviors are undertaken by consumers who are attempting to self-medicate feelings of anger, anxiety and paranoia. 3. We propose that it is likely at the juncture between the Sad and Angry positions on the emotional map that most compulsive consumption will occur: the consumer must feel unhappy, but still have the physical and mental energy (i.e., arousal) to get to the casino, liquor store, cigarette machine, shopping mall, or crack house in order to remedy his/her misery. Hirschman, Elizabeth C. and Morris B. Holbrook (1982), 'Hedonic Consumption: Emerging Concepts, Methods and Propositions.' Journal of Marketing, 46 (3), 92-101. Holbrook, Morris B. and Elizabeth C. Hirschman (1982),' The Experiential Aspects of Consumption: Consumer Fantasies, Feelings, and Fun.' Journal of Consumer Research, 9 (September), 132-140. Isen, A. M. (1984), Toward understanding the role of affect in cognition. In R. Wyer and T. Srull (eds.). Handbook of social cognition (Vol. 3, pp. 179-236). Hillsdale, NJ: Erlbaum. Isen, A. M. (1987), Positive affect, cognitive processes, and social behavior. In L. Berkowitz (ed.). Advances in experimental social psychology (Vol. 20, pp. 203-253). New York: Academic Press. Isen, A. M., Johnson, M. M. S., Mertz, E., and Robinson, G. F. (1985), The influence of positive affect on the unusualness of word associations. Journal of Personality and Social Psychology, 48,1413-1426. Isen, A. M., Rosenzweig, A. S., and Young, M. J. (1991), The influence of positive affect on clinical problem solving. Medical Decision Making, 11 (3), 221-227. Izard, Carroll E. (1977), Human Emotions, New York: Plenum.. Kahn, Barbara E. and Alice M. Isen, (1993), "The Influence of Positive Affect on Variety Seeking Among Safe, Enjoyable Products", Journal of Consumer Research, Vol. 20, (September), 2, 257-270. Larsen, R. J., and Diener, E. (1987a), A multitrait-multimethod examination of affect structure: Hedonic level and emotional intensity. Personality and Individual Differences, 6, 631-636. Larsen, R. J., and Diener, E. (1987b), Affect intensity as an individual difference characteristic: A review. Journal of Research in Personality, 21,1-39. Larsen, R. J., Diener, E., and Emmons, R. A. (1985), Affect intensity and reactions to daily life events. Journal of Personality and Social Psychology, 51, 803-814. Mednick, M. T., Mednick, S. A., and Mednick, E. V. (1964), Incubation of creative performance and specific associative priming. Journal ofAbnormal and Social Psychology, 69, 84-88. Mehrabian, Albert and James A. Russell (1974), An Approach to Environmental Psychology, Cambridge, MA: MIT Press. Murry, John P. and Peter A. Dacin (1996), 'Cognitive Moderators of Negative-Emotion Effects: Implications for Understanding Media Context.' Journal of Consumer Research, Vol. 22, March, 439-446. O'Guinn, Thomas C. and Ronald J. Faber (1989), 'Compulsive Buying: A Phenomenological Exploration.' Journal of Consumer Research, 16 (September), 147-157. Plutchik, Robert {19Wi), Emotion: A Psychoevolutionary Synthesis, New York: Harper and Row. Raju, P. S. (1980), 'Optimum Stimulation Level: Its Relationship lo Personality, Demographics and Exploratory Behavior.' Journal of Consumer Research, 7 (December), 272282. Rindfieisch, Ark, James E. Burroughs and Frank Denton (1997), 'Family Structure, Materialism and Compulsive Consumption.' Journal of Consumer Research, Vol. 23, March, 312325. Rook, D. W., and Fisher, R. J. (1995), Normative influences on impulsive buying. Journal of Consumer Research, 22, 296304. Rook, Dennis (1987), 'The Buying Impulse.' Journal of Consumer Research, 14 (September), 189-199.

REFERENCES
Aaker, David A. and Douglas M. Stayman (1989), 'What Mediates the Emotional Response to Advertising? The Case of Warmth.' In Cognitive and Affective Responses to Advertising, ed. Patricia Cafferata and Alice M. Tybout, Lexington, MA: Lexington. Allen, Chris T., Karen Machleit and Susan Schultz Kleine (1992), 'A Comparison of Attitudes and Emotions as Predictors of Behavior at Diverse Levels of Behavioral Experience.' Journal of Consumer Research, \o\. 18, March, 493-504. Amould, Eric J. and Linda L. Price (1993), 'River Magic: Extraordinary Experience and the Extended Service Encounter.' Journal of Consumer Research, Vol. 20, June, 24-45. Batra, Rajeev and Douglas M. Stayman (1990), 'The Role of Mood in Advertising Effectiveness.' Journal of Consumer Research, 17 (September), 203-214. Celsi, Richard L., Randall L. Rose and Thomas W. Leigh (1993), 'An Exploration of High Risk Leisure Consumption Through Skydiving.' Journal of Consumer Research, Vol. 20, June, 1-23. Edell, Julie and Marian C. Burke (1987), 'The Power of Feelings in Understanding Advertising Effects.' Journal of Consumer Research, 14 (December), 421-433. Estrada, Carlos A., Alice M. Isen, and Mark J. Young (1995), 'Positive Affect Facilitates Integration of Information and Decreases Anchoring in Reasoning Among Physicians.' Working Paper, Henry Ford Hospital and Cornell University. Faber, R. J. and O'Guinn, T. C. (1992), A clinical screener for compulsive buying. Journal of Consumer Research, 19,459469. Faber, R. J., Christenson, G. A., de Zwaan, M., and Mitchell, J. (1995), Two forms of compulsive consumption: Comorbidity of compulsive buying and binge eating. Journal of Consumer Research, 22, 296-304. Gardner, Meryl P. (1985), 'Mood States and Consumer Research: A Critical Review.' Journal of Consumer Research, 12 (December), 281-300. Goldberg, Marvin E. and Gerald J. Gom (1987), 'Happy and Sad TV Programs: How They Affect Reactions to Commercials.' Journal of Consumer Research, 14 (December), 387-403. Havlena, William J. and Morris B. Holbrook (1986), 'The Varieties of Consumption Experience: Comparing Two Typologies of Emotion in Consumer Behavior.' Journal of Consumer Research, 13 (December), 394-404. Hirschman, Elizabeth C. (1992), 'The Consciousness of Addiction: Toward a General Theory of Compulsive Consumption.' Journal of Consumer Research, 19 (September), 155-179.

Advances in Consumer Research (Volume 26) 111 Stayman, Etouglas M. and David A. Aaker (1988), 'Are all the Effects of Ad-Induced Feelings Mediated by Aad?' Journal of Consumer Research, Vol. 15, December, 368-373. Swinyard, William R. (1993), 'The Effects of Mood, Involvement, and Quality of Store Experience on Shopping Intentions.' Journal of Consumer Research, Vol. 20, September, 271-280.