The Role of Emotions in Marketing

JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 1999Bagozzi et al. / THE ROLE OF EMOTIONS

The Role of Emotions in Marketing

Richard P. BagozziUniversity of Michigan

Mahesh GopinathTulane University

Prashanth U. NyerChapman University

Emotions are mental states of readiness that arise from ap-praisals of events or one’s own thoughts. In this article, theauthors discuss the differentiation of emotions from affect,moods, and attitudes, and outline an appraisal theory ofemotions. Next, various measurement issues are consid-ered. This is followed by an analysis of the role of arousalin emotions. Emotions as markers, mediators, and mod-erators of consumer responses are then analyzed. Theauthors turn next to the influence of emotions on cognitiveprocesses, which is followed by a study of the implicationsof emotions for volitions, goal-directed behavior, and de-cisions to help. Emotions and customer satisfaction arebriefly explored, too. The article closes with a number ofquestions for future research.

This article addresses emotional behavior in marketing.In comparison to information processing and behavioraldecision research, we know much less about the role ofemotions in marketing behavior. Much of what we doknow is confined to consumer behavior, as opposed to thebehavior of salespeople or marketing managers.

Nevertheless, emotions are central to the actions ofconsumers and managers alike. Our goal in this article willbe to present a framework for thinking about emotions; todiscuss the measurement of emotions; to review how emo-tions function as causes, effects, mediators, and moderators

in marketing behavior; and to provide suggestions forfuture research.

THEORY AND FUNCTION OF EMOTIONS

Little consistency can be found in the use of terminol-ogy related to emotions. For purposes of organization anddiscussion, we begin with a definition of emotions andthen turn to a framework for interpreting emotionalbehavior.

Definitions

The termaffectwill be conceived herein as an umbrellafor a set of more specific mental processes including emo-tions, moods, and (possibly) attitudes. Thus, affect mightbe considered a general category for mental feelingprocesses, rather than a particular psychological process,per se.

By emotion, we mean a mental state of readiness thatarises from cognitive appraisals of events or thoughts; hasa phenomenological tone; is accompanied by physiologi-cal processes; is often expressed physically (e.g., in ges-tures, posture, facial features); and may result in specificactions to affirm or cope with the emotion, depending onits nature and meaning for the person having it. For a simi-lar perspective, see Lazarus (1991) and Oatley (1992).

The line between an emotion and mood is frequentlydifficult to draw but often by convention involves conceiv-ing of a mood as being longer lasting (from a few hours upto days) and lower in intensity than an emotion. Yet, excep-tions to this construal can be found. Still anotherdistinction

Journal of the Academy of Marketing Science.Volume 27, No. 2, pages 184-206.Copyright © 1999 by Academy of Marketing Science.

between emotions and moods is that the former typically isintentional (i.e., it has an object or referent), whereasmoods are generally nonintentional and global or diffused(Frijda 1993). Also, moods are not as directly coupled withaction tendencies and explicit actions as are manyemotions.

Finally, attitudes, too, are often considered instances ofaffect, with the same measures used on occasion to indicateemotions and attitudes (e.g., pleasant-unpleasant, happy-sad, or interested-bored semantic differential items). How-ever, some authors take a narrower view of attitudes anddefine them as evaluative judgments (measured, e.g., bygood-bad reactions) rather than emotional states. Cohenand Areni (1991), for instance, reserve the termaffectfor“valenced feeling states,” with emotions and moods asspecific examples. Attitudes are evaluative judgments intheir view. Nevertheless, other researchers do not make adistinction between affect and evaluative judgments. Forexample, Eagly and Chaiken (1993) point out that Fish-bein and Ajzen (1975) and other social psychologists have“regarded affect as isomorphic with evaluation itself andused the terms interchangeably” (p. 12). Still others pro-pose that attitudes have two distinct, but generally highlycorrelated, components: affective and cognitive (orevaluative) dimensions. Some empirical support exists forthis interpretation (Bagozzi and Burnkrant 1979; Batraand Ahtola 1990; Breckler and Wiggins 1989; Crites, Fab-rigar, and Petty 1994; Eagly, Mladinic, and Otto 1994).

It should be recognized that the termsaffect, emotions,moods, andattitudeshave frequently been used inconsis-tently in the literature. We will revisit this issue when weconsider both measurement issues and customer satisfac-tion research below. For now, we stress that when readingthe literature, it is important to pay attention to howauthors define affective (and related) terminologies andhow they measure the variables to which the terminologiesrefer. One’s definition of terms permits an interpretation oftheir meaning, but equally important is how the variablesto which the terms refer are operationalized. Some authorshave defined key variables as emotions, moods, or atti-tudes but have used operationalizations corresponding todifferent concepts. Other authors have used operationali-zations for a single variable that cut across two or moreinstances of affect. To make clear our definition of emo-tions and how it differs from definitions of mood and atti-tudes, we present the following point of view.

Organizing Framework

Above we noted that emotions are mental states ofreadiness. But so, too, are moods and attitudes. How thenmight we distinguish between these affective states? Forone thing, the state of readiness characterized by an emo-tion tends to be more intense than that characterized bymoods or attitudes. It is more intense in the sense of

strength of felt subjective experience, plus magnitude ofphysiological response (e.g., autonomic nervous systemactivity) and extent of bodily expression (e.g., facial dis-plays), when these latter reactions accompany an emotion.

Probably the most important factor differentiatingemotions from moods and attitudes is the way emotionsarise. Emotions are said to have a specific referent (e.g., aconsumer becomes pleased when a new detergent removesgrass stains from clothing; he or she is angered by poor ser-vice in a restaurant). Specifically, emotions arise inresponse to appraisals one makes for something of rele-vance to one’s well-being. By appraisal, we mean anevaluative judgment and interpretation thereof. By some-thing of relevance, we mean an incident or episode thathappens to oneself (e.g., an unplanned event); a behaviorone performs or a result one produces (e.g., engaging in anactivity or receiving or failing to receive a planned out-come); or a change in an object, person, or thought that haspersonal meaning.

It is important to stress that although categories ofevents or physical circumstances are frequently associatedwith particular emotional responses, it is not the specificevents or physical circumstances that produce the emo-tions but rather the unique psychological appraisal madeby the person evaluating and interpreting the events andcircumstances. Different people can have different emo-tional reactions (or no emotional reactions at all) to thesame event or happening. Note, too, that appraisals can bedeliberative, purposive, and conscious, but also unreflec-tive, automatic, and unconscious, depending on the personand eliciting conditions for emotional arousal. The centralrole of appraisals in the formation of emotions has come todefine what are aptly called appraisal theories in psychol-ogy (e.g., Frijda 1986; Lazarus 1991; Ortony, Clore, andCollins 1988; Roseman 1991; Smith and Ellsworth 1985).

Appraisal theorists maintain that the critical determi-nant of any emotion is the resultant evaluation and inter-pretation that arise after comparing an actual state with adesired state. Two appraisals are particularly crucial at thisstage of emotion formation: goal relevance and goal con-gruence (Lazarus 1991). That is, a necessary condition foran emotional response to an event or happening is that aperson has a personal stake in it and at the same timejudges the event or happening to facilitate or thwart thisstake. Again, the appraisal can occur consciously orunconsciously.

A distinctive feature of appraisal theories is their speci-fication of the conditions leading to discrete emotionalresponses. Forced to be brief, we focus on Roseman’s(1991) version of appraisal theories, which differs in rela-tively minor ways from other leading theories. Rosemanhypothesized that particular combinations of five apprais-als determine which of 16 unique emotions will be experi-enced in any given situation. Figure 1 summarizes his the-ory, where the five appraisals are labeled motive

Bagozzi et al. / THE ROLE OF EMOTIONS 185

consistent/motive inconsistent (i.e., positive emotions ver-sus negative emotions), appetitive/aversive (i.e., presenceof a reward vs. absence of a punishment), agency (i.e., out-come is perceived caused by impersonal circumstances,some other person, or the self), probability (i.e., an out-come is certain or uncertain), and power (i.e., strong ver-sus weak coping potential).

For example, pride occurs when one evaluates his or herown performance of an action or achievement of an out-come in a positive light (e.g., a feeling of having donewell). Here the positive emotion is motive consistent,either appetitive (e.g., having attained a positive goal) oraversive (e.g., having avoided a punishment), self-produced under weak or low coping potential, and eithercertain or uncertain, depending on the circumstances. Sad-ness happens when one experiences a loss for which onerecognizes that nothing can be done to restore it. The loss,which is of something or someone valued, is experiencednegatively and with high certainty under conditions of lowcoping power. It is perceived to be caused by impersonalcircumstances.

One value of appraisal theories is that it is possible toaccount for most emotions. Indeed, subtle combinations ofappraisals yield discrete emotional responses. Anger andregret, for example, differ primarily in only one type ofappraisal and share in the other four, namely, anger occurswhen a person sees another person as the source of injuryto oneself or to another person viewed as a victim of injus-tice, whereas regret results when one’s negative outcomeis attributed to actions or inactions of the self.

Not every emotion is accounted for by Roseman’sframework (or by any other framework for that matter).For instance, pride is regarded as a positive emotion inRoseman’s framework, yet excessive or exaggerated pride(sometimes termedhubris) can invite retribution. Like-wise, shame and guilt are thought by Roseman to be pro-duced by similar appraisals, but other researchers havefound important distinctions between shame, guilt, andembarrassment (e.g., Lewis 1993). Likewise, disgust hasbeen studied extensively and found to differ from distress(e.g., Rozin, Haidt, and McCauley 1993). Nevertheless, incontrast to other theories of emotion that conceive of it inbipolar terms (e.g., pleasure-displeasure and high arousal-low arousal [Russell 1980] or high negative affect-lownegative affect and high positive affect-low positive affect[Watson and Tellegen 1985]), Roseman’s framework andother appraisal theories not only allow for many discreteemotions but specify conditions for their occurrence.

An elaboration of appraisal theories that is especiallyrelevant for marketing is the treatment of goals, which maybe defined as “internal representations of desired states,where states are broadly defined as outcomes, events, orprocesses” (Austin and Vancouver 1996:338). Oatley andJohnson-Laird (1987) proposed what they termed acom-municative theory of emotionswherein events are evalu-ated in relation to a person’s goals. Emotions are thoughtto function to coordinate parts of one’s cognitive system soas to manage responses to events and in so doing changefrom ongoing to new activities or to maintain desired statesor activities. The self-regulation of goals is believed to bethe main function of emotions:

Each goal and plan has a monitoring mechanism thatevaluates events relevant to it. When a substantialchange in probability occurs of achieving an impor-tant goal or subgoal, the monitoring mechanismbroadcasts to the whole cognitive system a signalthat can set it into readiness to respond to thischange. Humans experience these signals and thestates of readiness they induce as emotions. (Oatley1992:50)

According to Oatley and Johnson-Laird (1987), emotionsare evoked “at a significant juncture of a plan . . . typi-cally . . . when the evaluation (conscious or unconscious)of the likely success of a plan changes” (p. 35). Positiveemotions (e.g., happiness, elation, joy) are associated withthe attainment of a (sub)goal, which usually leads to a de-cision to continue with the plan, whereas negative emo-tions (e.g., frustration, disappointment, anxiety) resultfrom problems with ongoing plans and failures to achievedesired goals (see also Stein, Liwag, and Wade 1996).

Emotions have implications for action and goal attain-ment. Lazarus (1991) identifies coping responses asimportant mechanisms in this regard. When we experience

186 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 1999

FIGURE 1Roseman’s (1991) Appraisal

Theory of Emotions

SOURCE: Roseman (1991:193). Reprinted with permission.

a negative emotion (e.g., anger, sadness, fear), we are indisequilibrium and wish to return to our normal state.Either one or both of two coping processes are typicallyused: problem-focused coping, where we attempt to alle-viate the sources of distress, or emotion-focused coping,where we either change the meaning of the source of dis-tress (e.g., deny that a threat exists, distance oneself fromthe source of distress) or avoid thinking about a problem.

By contrast, coping with positive emotions ofteninvolves sharing one’s good fortune, savoring the experi-ence, working to continue or increase the rewards, andincreasing physical activity. Positive emotions are some-times accompanied as well by higher levels of physiologi-cal arousal, expanded attention, increased optimism,enhanced recall, and a shift from self- to other-centeredorientations (e.g., becoming friendlier, caring about oth-ers), when compared, say, to sadness. Indeed, positiveemotions, particularly happiness, frequently stimulatehelping or altruistic actions. Why? Schaller and Cialdini(1990) offer two explanations: “First, we may propose thatpositive mood leads to enhanced helping via the morepositive outlook and enhanced activity that appear tospring automatically from the experience of happiness,”and second, “we argue that happiness is associated with amotivation toward disequilibrium—toward the possibleattainment of additional personal rewards that transcendthe basic concern over one’s mood” (pp. 284-285). Thepersonal rewards referred to here concern such self-enrichment motives as affiliation, achievement, compe-tence, and esteem.

Closely related to coping responses are action tenden-cies. An action tendency is “a readiness to engage in or dis-engage from interaction with some goal object” andincludes “(i)mpulses of ‘moving towards,’ ‘moving away,’and ‘moving against’ ” (Frijda, Kuipers, and ter Schure1989:213). Some theorists maintain that emotions are notmerely reactions to appraisals of events but also includeaction tendencies as part of their meaning (Frijda 1986).Others go further and maintain that action tendencies areautomatic, “prewired” responses connected to emotions(LeDoux 1996). And in Frijda’s (1986) treatment, emo-tions are conceived as the entire process from stimulusevent to action and arousal:

Stimulus event→event coding→appraisal (evaluation of

relevance, context, and urgency/difficulty/seriousness of

event)→action readinessaction

arousal

Much as emotions arise in response to patterns of apprais-als, Frijda (1986; Frijda et al. 1989) has shown that pat-terns of action readiness correspond to distinct emotioncategories (e.g., avoidance with fear, helping with caring,

helplessness with sadness, assault with anger, withdrawalwith shame).

Finally, it has been argued that many coping responsesto emotions are volitional (Bagozzi 1992:186-189). Theprocess begins with outcome-desire units and appraisalsof changes or anticipated changes in goal attainment orgoal progress. Four appraisal classes can be identified.Outcome-desire conflicts happen when one fails toachieve a goal or when one experiences an unpleasantevent. One or more emotional reactions occur to outcome-desire conflicts (e.g., dissatisfaction, anger, shame, guilt,sadness, disappointment, disgust, regret), depending onattributions of the source of goal failure or the unpleasantevent (i.e., self, other person, or external cause). The cop-ing response(s) to these emotions, in turn, is selected fromthe following: intent to remove or undo harm, obtain helpor support, decrease outcome, reevaluate goal, or redoubleeffort, if appropriate, depending on the specific emotioninvolved.

Outcome-desire fulfillment takes place when oneachieves a goal, experiences a pleasant event, or avoids anunpleasant event. One or more emotional reactions comeabout when outcome-desire fulfillment happens (e.g., sat-isfaction, joy, elation, pleasure, pride, relief, caring, love),again depending on attributions of the source of good for-tune. The coping responses to these emotions include anintention to maintain, to increase, to share, or to enjoy theoutcome.

Outcome-desire conflicts and fulfillment refer to out-comes in the past or present. The following two appraisalclasses go on with regard to planned outcomes. Outcome-desire avoidances transpire in anticipation of unpleasantoutcomes or goals. Fear or its variants (e.g., worry, anxiety,distress) are the emotional reactions to this appraisal. Thecoping responses to these emotions entail either an intentionto avoid undesirable outcomes or to reinterpret the threat.

The final class of appraisals, outcome-desire pursuits,happen in anticipation of pleasant goals or outcomes.Hope is the emotional reaction to such appraisals. Thecoping response(s) to hope includes intentions to realize orfacilitate outcome attainment and to sustain one’s commit-ment and vigilance.

The theory of self-regulation suggests that unique voli-tional responses underly coping for each particular emo-tion or class of emotions (Bagozzi 1992). In addition, thespecific intention enacted depends on one’s degree of self-efficacy in executing the coping responses. Somewhatanalogous (appraisal→emotional reactions→coping)responses occur for outcome-identity conflicts, fulfill-ments, avoidances, and pursuits in social situations relatedto normative expectations (Bagozzi 1992:191-194).

Returning to the distinction between emotions andmoods and attitudes, we might say that in addition to thethings mentioned earlier, emotions differ from moods andattitudes in the manner in which they arise and in their


representation in memory. Emotions occur in response tochanges in specific plans or goal-relevant events. As Oat-ley (1992) points out, emotions are manifest as “transi-tions from one sequence of action and another,” but moodsoccur “when the cognitive system is maintained in an emo-tion mode for a period” (pp. 64, 91-92). Indeed, moods areoften resistant to changes in events surrounding them. Onereason for this is that moods “depend on the dissociabilityof control emotion signals from semantic informationabout causation” (Oatley 1992:64). In general, moods areelicited by “(a) after effects of emotions; (b) organismicconditions such as illness, fatigue, previous exercise, andgood health, or pharmacological agents; (c) general envi-ronmental conditions and side-effects of activities: heat,noise, environmental variety, stressful conditions” (Frijda1986:289).

Like emotions, attitudes can arise from changes inevents, but attitudes also occur in response to mundaneobjects. In addition, arousal is a necessary part of emotionsbut not necessarily attitudes. Moreover, attitudes seem tohave the capacity to be stored during long periods of timeand retrieved, whereas emotions are not experienced inthis way (i.e., emotions are ongoing states of readiness;they are not stored and retrieved, per se, although it is pos-sible to recreate the conditions originally producing themin our memory and react emotionally to the thoughts sogenerated, at least up to a point; emotions can, however, beclassically conditioned, but it is unclear whether attitudescan). Finally, the connection of emotions to volition andaction is stronger and more direct than it is for attitudes.Emotions directly stimulate volitions and initiate action,but attitudes may require an additional motivation impe-tus, such as desire (Bagozzi 1992).

MEASUREMENT OF EMOTIONS

The measurement of emotions could focus on a full setof signs or evidence, including evaluative appraisals, sub-jective feelings, body posture and gestures, facial expres-sions, physiological responses, action tendencies, andovert actions. Whatever measurements one uses should, ofcourse, be tied to an underlying theory of emotions.

Some authors (e.g., most appraisal theorists) construeemotions as mental states or processes, and thus it wouldbe prudent to directly measure the cognitive activitiescomprising the emotional content of these states orprocesses, from this perspective. Self-reports of one’s sub-jective experiences constitute the most frequently usedprocedures in this regard, although other methods for indi-cating emotional memory processes might be used as well(e.g., response time, subliminal priming). From the pointof view of mental conceptualizations of emotions, physio-logical, motor, or biological indicators would be at bestconsidered correlates or indirect measures of emotions.

Other researchers who interpret emotions in broaderterms, as either the whole process from the coding ofevents to action responses (e.g., Frijda 1986) or as com-plex patterns of physiological responses (e.g., Cacioppo,Berntson, and Klein 1992 maintain that emotional experi-ence is a function of somatovisceral activation, afferentia-tion, and cognitive operations; LeDoux 1996 emphasizesbrain processes, especially the role of the amygdala) stressthe need for measurement processes going beyond self-reports. Depending on the theorist, overt behaviors orphysiological reactions may be considered either a part ofwhat it means to have an emotion or antecedents, concomi-tants, or possibly even effects of an emotion. More behav-iorally or physiologically oriented researchers obviouslyemploy measures of emotions consistent with theseinterpretations.

Marketers have tended to take an empirical approach tothe measurement of emotions and to rely on self-reports(i.e., either unipolar or bipolar items on questionnaires). Inthe typical application, many items cutting across numer-ous positive and negative emotions are administered tomeasure reactions to a stimulus, and such methods as factoranalysis, multidimensional scaling, or cluster analysis areused to identify the underlying emotional dimensions forthe sample at hand. The number of items investigated inthis regard has been as large as 180 (Aaker, Stayman, andVezina 1988), while a paper-and-pencil technique (basi-cally a single item measured continuously while viewingan ad) has even been suggested to register “warmth”toward a stimulus ad (Aaker, Stayman, and Hagerty 1986;see also Russell, Weiss, and Mendelsohn 1989).

Two influential studies in the measurement of emo-tional responses toward advertisements are those by Edelland Burke (1987) and Holbrook and Batra (1987). Edelland Burke (see also Burke and Edell 1989) developed a 52-item scale for measuring emotions towards ads, while Hol-brook and Batra worked with a 94-item scale, which waslater reduced to 34 items (Batra and Holbrook 1990).

With so many items measuring emotions, a questionarises whether a small number of basic dimensions under-lie people’s responses. Edell and Burke (1987) analyzedthe items in their scale and found three factors: upbeat feel-ings, negative feelings, and warm feelings. Likewise, Hol-brook and Batra (1987) used factor analysis, but in aslightly different way. Their 94 items were first generateda priori to measure 29 emotional indices. For example,joyful, happy, delighted, and pleased were hypothesized toindicate a joy index, and ashamed, embarrassed, andhumiliated were hypothesized to indicate a shame index.Then, based on factor analyses of the 29 indices, a three-factor solution for emotions was found: pleasure, arousal,and domination. Similarly, Batra and Holbrook (1990)factor analyzed 12 emotional indices (largely a subset ofthose revealed in Holbrook and Batra, with a few excep-tions) and discovered three factors corresponding closely


to those discovered by Edell and Burke. Other researchers(e.g., Oliver 1994; Westbrook 1987), also using factoranalysis, have found emotional items to load on two fac-tors: positive affect and negative affect.

Richins (1997) recently argued that consumption-related emotions are more complex than the two- andthree-factor solutions observed in studies of reactions toads or customer satisfaction. Moreover, because explora-tory factor analyses often yield a small number of factors,she used a multidimensional scaling procedure, in con-junction with examination of clusters based on locationand semantic similarity of emotional descriptors in two-dimensional space. Sixteen clusters of emotions wereidentified, each measured by 2 to 8 indicators (in Study 4):anger, discontent, worry, sadness, fear, shame, envy, lone-liness, romantic love, love, peacefulness, contentment,optimism, joy, excitement, and surprise.

Although the approaches used by marketers to datehave been largely empirically driven (e.g., Edell and Burke1987; Holbrook and Batra 1987; Oliver 1994; Richins,1997; Westbrook 1987), they are consistent in certain senseswith leading perspectives on emotions in psychology. Forexample, Holbrook and Batra’s (1987) three-factorpleasure-arousal-domination findings are somewhat simi-lar to Russell and Mehrabian’s (1977) three-factorpleasure-arousal-dominance model. However, some dif-ferences can be pointed out. The most important are thehigh loadings of sadness and fear on the domination factorin Holbrook and Batra’s study, in contrast to the more com-mon outcome of sadness loading on or near the negativepole of a pleasure-displeasure factor, and fear loadingabout 45 degrees away from displeasure and towardgreater arousal (e.g., Russell, 1997).

Similarly, the three factors found in Edell and Burke’s(1987) study (i.e., upbeat feelings, negative feelings, andwarm feelings), correspond roughly to the high positiveaffect-low positive affect, high negative affect-low nega-tive affect, and pleasantness-unpleasantness dimensions,respectively, of Watson and Tellegen’s (1985) circumplexmodel; the positive-affect and negative-affect factorsobserved by Oliver (1994) and Westbrook (1987) alsoalign to a great extent with factors on the circumplexmodel (see also Mano and Oliver 1993).

The circumplex structure of emotions is shown in Fig-ure 2. This representation of emotions has also been calledthe two-factor model, because, based on the techniquesused to generate it (e.g., factor analysis or multidimen-sional scaling), emotions can be arranged around twoorthogonal axes. Russell (1997) terms the axes pleasure-displeasure and arousal-sleepiness, while Watson and Tel-legen (1985) label them high positive affect-low positiveaffect and high negative affect-low negative affect. Rus-sell’s interpretation is essentially the same as Watson andTellegen’s pleasantness-unpleasantness and strongengagement-disengagement axes, which are rotated 45

degrees from their primary axes (but see Larsen andDiener 1992).

The idea behind the circumplex model is that emotionsexist in bipolar categories (e.g., happy-sad, nervous-relaxed) and can be arranged in a continuous order aroundthe perimeter of a two-factor space. The closer emotionsare to each other on the perimeter, the more similar theyare. For example, excited and aroused are more similarthan are content and aroused (see Figure 2). The origin orcenter of the circumplex is thought to represent a neutralpoint or adaptation level.

The circumplex model is appealing because it is intui-tive, simple, and provides a description of which emotionsare similar and which are dissimilar. However, it has draw-backs. The most serious limitation is that it is based onempirical associations among experienced emotions andhas nothing to say about the conditions (e.g., appraisals)producing emotions. Then, too, the circumplex model canobscure subtle differences in emotions. Depending oneliciting conditions and people’s appraisals, each of theemotions grouped together within any particular categoryon the circumplex can be distinct from its cocategorymembers. For example, it is possible to feel fearful withoutfeeling hostile (see “high negative affect” in Figure 2).Then, too, the circumplex contains categories that may notcorrespond to emotions. For instance, surprised, drowsy,and sleepy do not seem to reflect emotions. On the other


FIGURE 2Watson and Tellegen’s Two-Factor

Structure of Affect

SOURCE: Watson and Tellegen (1985:225). Copyright © 1985 by theAmerican Psychological Association. Reprinted with permission.

hand, the circumplex fails to represent well instances ofemotion important in everyday life and marketing. It does notaccommodate love, disgust, pride, hope, guilt, shame, orembarrassment very well, to name a few. The various two-and three-factor summaries of emotions disclosed in mar-keting also include variables not reflective of emotions andexclude instances generally recognized as emotions.

Richins’s (1997) Consumption Emotions Set (CES)with its 16 descriptors is appealing because it covers mostemotional reactions one encounters in consumption, andits measures achieved satisfactory reliability (except formeasure of envy, loneliness, peacefulness, and content-ment). The CES would be best used within the context of aparticular theory of emotions to operationalize specificcategories of emotions hypothesized to serve as antece-dents, consequences, or moderating variables. The CEScould also be used to operationalize emotions in moreempirically oriented studies, but whether measures of eachdimension would achieve discriminant validity is prob-lematic. Most studies incorporating multiple instances ofboth positive emotions and negative emotions find that themeasures load on two factors corresponding to positiveand negative emotions (e.g., Bagozzi, Baumgartner, andPieters, 1998; Oliver, 1994).

This raises the question of when one can expect discreteemotional reactions versus amalgamated groupings of,say, positive emotions and negative emotions (e.g., highlycorrelated feelings of anger, sadness, and fear). The advan-tage of a theory-based approach to emotions is that spe-cific conditions can be specified for the occurrence of dis-tinct emotions, and these hypotheses can be tested. It is forthese reasons that we used appraisal theories as our organ-izing framework for looking at emotions. Discrete emo-tional reactions are likely to happen when one manipulatesconditions producing specific appraisals or when naturallyoccurring events correspond to unique appraisal condi-tions. Gopinath and Bagozzi (1999), for example, wereable to induce independent emotional reactions towardthree targets in a moviegoing context. On the basis ofRoseman’s (1991) theory, distinct emotional reactionswere produced as a function of three-way interactionsbetween motive consistency-inconsistency, appetitive-aversive, and self-other agency conditions. For example,admiration, affection, dislike, and contempt resultedtoward a group member in decision making with regard tomovie choice, and pride, shame/guilt, and regret resultedtoward the self in group decision making, depending onthe three appraisal conditions. Likewise, happiness, satis-faction, annoyance, or frustration occurred toward themovie, and pleasure, contentment, irritation, or angeroccurred toward the theater, depending on the threeappraisal conditions. Scenarios were used to create theappraisal conditions.

By contrast, research not based on manipulations ofappraisal conditions or based on reactions to a singlestimulus frequently finds that emotions cluster in two and,on occasion, three factors (e.g., Edell and Burke 1987;Holbrook and Batra 1987; Oliver 1994; Westbrook 1987).Furthermore, research examining the construct validity ofmeasures of discrete emotions obtained in nonexperimen-tal survey settings shows that discriminant validity is oftenlacking among measures of different positive or differentnegative emotions (e.g., Bagozzi 1993) or between mea-sures within a particular subcategory of positive or negativeemotions, such as among measures of elation, gladness,and joy (Bagozzi 1991a).

What accounts for the differences in findings betweenexperimental research based on appraisal theories and sur-vey research or research based on reactions to a singlestimulus measured by inventories of emotional items? Onepossibility may be that discrete emotional reactions areshort-lived or, once activated, stimulate other emotionalreactions closely related to them. Consider, for example,sadness: “[W]hen we experience loss, we rarely feel a sin-gle emotion such as sadness. We grieve, are angry, anx-ious, guilty, envious, even hopeful, and defensive” (Lazarus1991:250). A reason why these emotions may go in tan-dem is that coping processes for sadness may involveactive struggle or even protest against loss, which results inother emotions. Alternatively, the absence of discreteemotions may simply reflect how difficult it is to createthem. As Izard (1972) noted, pure emotions are “virtuallyimpossible to obtain in the laboratory or in any researchsetting” (p. 103). However, as noted above, researchershave recently found that the use of scenarios in an experi-mental context can generate discrete emotionalresponses(e.g., Gopinath and Bagozzi 1999; Roseman 1991).

Another factor that might account for a coalescence ofmultiple positive emotions and multiple negative emotionsin two corresponding groupings is the nature of the stimu-lus under study. Most stimulus ads, products, or brands arecomplex, and the appraisals engendered are typicallyvariegated, but related. Also the way in which items arepresented on some questionnaires makes it difficult touncover discrete emotional components. When multiplemeasures of a single discrete emotional response (e.g.,happy, pleased, and joyful for “joy”) are interspersedthroughout a questionnaire, this tends to reduce correla-tions among items purported to indicate the same responseand to increase correlations of these items with measuresof other responses. The result is predictably a reduction indiscriminant validity and high correlations among itemsmeasuring positive emotions and among items measuringnegative emotions. The alternative is to group items by theemotional response they are intended to tap, which tendsto increase correlations among measures of the same thing


and decrease correlations among measures of differentemotional responses. Thus, a trade-off is entailed by use ofeither practice.

An issue that has received little attention in marketing iswhether to use unipolar or bipolar items to measure emo-tions. The choice can influence findings and their interpre-tation in fundamental ways. Some leading scholars claimthat emotions are in the final analysis bipolar states orprocesses. We are either happy or sad, for example, andany other pattern (e.g., independence or concomitance) isthought to be an artifact of measurement error. Although anumber of studies have shown that pleasant and unpleasantemotions are independent (e.g., Bradburn 1969; Dienerand Emmons 1985; Zevon and Tellegen 1982), Green,Goldman, and Salovey (1993) and Barrett and Russell(1998) argue and present findings demonstrating that emo-tions are bipolar, once random or both random and system-atic errors are taken into account.

Bagozzi, Wong, and Yi (1998) challenge the conclu-sions made by Green et al. (1993) and Barrett and Russell(1998). They hypothesize that bipolarity, independence,and concomitance depend on gender, culture, and the tar-get of one’s emotions. Briefly, Bagozzi, Wong, and Yifound that positive and negative emotions were highlynegatively correlated for American women but highlypositively correlated for Chinese women. The former pat-tern is evidence for bipolarity (i.e., either positive or nega-tive emotions occur but not both), the latter for concomi-tance (i.e., both positive and negative emotions occur at thesame time). For men, the correlations between positive andnegative emotions were much smaller in magnitude but inthe same direction across ethnicity, that is, slight negativecorrelations were found for American men, slight positivecorrelations for Chinese men. Thus, the relationshipbetween positive and negative emotions for men wasnearly independent. Bagozzi, Wong, and Yi argued thatdifferences in culture (i.e., a tendency for Americans toview things in dichotomies or discrete categories, i.e., inopposition; and a tendency for Chinese to view things dia-lectically, i.e., in balance or harmony) interact with genderdifferences (i.e., a tendency for women to be more knowl-edgeable and skilled in the use of emotions than men) toproduce the divergent patterns. The above findingsresulted when people were asked to express how they feltat the moment according to their idiosyncratic reasons,which is the standard procedure used in the literature. Thatis, the stimulus for each person could be considered het-erogeneous across individuals. However, when Americansand Chinese were asked to give their emotional reactionsto eating in fast-food restaurants (a common, singularstimulus), positive and negative emotions were indepen-dent for men and women alike.

It is perhaps too early to give definitive recommenda-tions on which emotional scales to employ in empiricalwork, but for now, we think that it is advisable to recom-mend use of unipolar scales that ask respondents toexpress to what extent each emotion describes their ownsubjective feelings, rather than bipolar scales that canobscure differences in emotional responses across thevarious dimensions. Also, at least five, preferably seven tonine, scale steps should be used for each item to enhancethe chances that optimal distributional properties of mea-sures will be achieved. In addition, at least three, prefera-bly more, items should be used for each emotionalsubcategory.

A final measurement issue we wish to raise is the fol-lowing. To what extent are emotions blends of categories?For ease of discussion, we consider the categories of emo-tions presented on the circumplex. Pleasantness-unpleasantness might combine, for instance, with arousalto produce different kinds or intensities of emotion. To takea particular example, consider happiness, an instant ofpleasantness on the circumplex. Intense forms of happinessoccur when pleasantness combines with high arousal:elated, excited, enthusiastic, euphoric, gleeful, joyous, ec-static, and exultant are examples. Mild forms of happinessoccur when pleasantness combines with low arousal:peaceful, calm, serene, or quietude of mind are examples.Happiness, itself, might be at an intermediate level ofarousal. Other perspectives on blended emotions, basednot on a blend of arousal with emotion categories butrather on combinations of “basic” emotions, can be foundin Izard (1991, 1992) and Plutchik (1980).

What more general role, if any, does arousal play inemotions (Bagozzi 1991b)? We turn to this issue next.

AROUSAL

An early, influential point of view on emotions was pro-fessed by James ([1890] 1950) who claimed that “bodilychanges follow directly the perception of the exciting fact,and . . . ourfeeling of the same changes as they occur IS theemotion” (p. 449, emphasis in original). For James, differ-ent stimuli lead to different bodily responses (e.g., sweatypalms, racing heart, etc.), these physiological responsesare then detected as bodily sensations in our mind, and theresult is interpreted by us as emotional experiences. But itis important to note that James reserved this interpretationfor what he termed the “coarser” emotions (e.g., “grief, fear,rage, love”), which involve strong bodily perturbations; hewas less clear about what he termed the “subtler” emotions(e.g., “moral, intellectual, and aesthetic feelings” (James[1890] 1950:468).

An equally influential theory of emotions was pro-posed by Schachter and Singer (1962), who argued that


emotion is essentially bodily arousal plus a cognitive labelone provides to diagnose his or her felt arousal, a perspec-tive consistent with James’s point of view. The idea is thatwe first experience physiological arousal, and, especiallywhen we are unaware or uncertain of the origin of thearousal, we look for evidence in the physical and socialsituation accompanying the arousal to label our emotionalstate. Although this theory had considerable impact in psy-chology for nearly three decades, it has largely been dis-credited and has not received much supporting evidencebeyond Schachter and Singer’s original experiment (e.g.,Manstead and Wagner 1981; Reisenzein 1983). Oneexception to the above observation is the frequentlyrepeated finding that arousal misattributed to an extrane-ous source intensifies emotions (e.g., Cantor, Bryant, andZillman 1974; Dutton and Aron 1974; Zillman 1971).

An important contribution of James ([1890] 1950) andSchachter and Singer (1962) was the recognition thatarousal plays an essential role in emotion. Before weaddress arousal more fully, we should mention research inmarketing that addresses the acquisition of affect, withoutnecessarily involving concepts of arousal or informationprocessing. Both in the practice of marketing and market-ing research, considerable emphasis has been placed onthe effects of various stimuli on consumer behavior. Retailstore environment cues, advertising, background music,brand names, packages, celebrity endorsers, and otherstimuli are frequently administered to produce emotionalreactions in consumers. The premise is that emotions ormoods trigger buying responses (e.g., Gardner 1985; Hilland Gardner 1987).

If not by appraisal processes or direct arousal, per se,how does presentation of a stimulus under repetitive con-ditions induce affect? A seductively simple explanationwas provided by Zajonc, who argued that “when objectsare presented to the individual on repeated occasions, themere exposure is capable of making the individual’s atti-tude toward these objects more positive” (Zajonc and Mar-kus 1982:125). This “mere exposure” effect has beenfound primarily when the stimulus is simple and previ-ously unknown or else has little or no semantic content(e.g., nonsense syllables, foreign words of Chinese char-acters). One mechanism that has been offered to explainthe mere exposure effect is familiarity: we come to likethings that are familiar to us, perhaps because of feelingsof security (Zajonc 1968). However, a full theoreticalexplanation for the mere exposure effect has not beendeveloped. When a person is exposed to meaningful stim-uli, it has been more difficult to produce the mere exposureeffect (e.g., Obermiller 1985). This is, in part, a conse-quence of the cognitive processing that occurs in responseto awareness of the meaningful stimuli. Repeated expo-sure to a meaningful stimulus can lead to increased ordecreased positive or negative feelings, depending on its

meaning. Mere exposure, thus, loses its utility in suchcases, both as a theory and a practical tactic.Based loosely on the notion that emotions or moodsinduced by one stimulus become attached to another, someresearchers have investigated the effects of (a) music onlength of stay and money spent in supermarkets and res-taurants (e.g., Milliman 1982, 1986) and (b) affective toneof stores on purchase intentions (e.g., Donovan and Rossi-ter 1982) and evaluations (e.g., Gardner and Simokos1986). How can the observed attachment of affect fromone stimulus to another be explained? Shimp (1991)reviews seven studies in consumer research that test vari-ous facets of classical conditioning explanations. The ideabehind classical conditioning is that the repeated pairingof a conditioned stimulus (e.g., a new brand name) with anunconditioned stimulus (e.g., an attractive spokesperson)will eventually lead to the new brand name, on its own,stimulating the unconditioned response (e.g., positiveaffect) originally induced by the unconditioned stimulus.Very few studies have been performed in marketing thatconform to the conditions required to test classical condi-tioning. It is unclear whether classical conditioning studiescan be designed to rule out such rival explanations asdemand characteristics or cognitive interpretations of theresults. Allen and Janiszewski (1989) provide some evi-dence that at least one type of cognitive mediation is neces-sary for classical conditioning to occur: namely, subjectawareness of the contingency between the conditionedstimulus and unconditioned stimulus. On the other hand,classical conditioning, particularly for fear responses, hasbeen shown to involve unconscious arousal processes con-nected with the amygdala (LeDoux 1996).

Another way to explain the observed attachment ofaffect from one stimulus to another is by Zillman’s (1971)excitation-transfer model. Briefly, Zillman proposed thatexposure to one stimulus may produce arousal. If a secondstimulus is presented close on that also is capable of pro-ducing arousal on its own, the two sources of arousal maycombine to produce intensely experienced arousal. Undercertain conditions (e.g., unawareness of the source ofarousal from the first stimulus, recency of the second), aperson may attribute the arousal to the second stimulus.Zillman (1983) interpreted arousal as undifferentiatedsympathetic activation.

An issue in need of resolution is whether emotions canoccur without arousal. Cognitive theories of emotions(e.g., appraisal theories) seem to allow that emotions canbe produced by cognition alone, without arousal (e.g., Par-rott 1988). But does arousal always accompany the experi-ence of emotions?

Recent research suggests that arousal is an essentialcomponent of emotion and is manifest in neural systems inthe brain. LeDoux (1996) reviews evidence suggestingthat there are at least five arousal systems in the brain con-tributing complexly to emotional experience. Four of


these are in regions of the brain stem and rely respectivelyon acetylcholine, noradrenaline, dopamine, and serotoninfor activation. A fifth (the nucleus basalis) is in the fore-brain and also relies on acetylcholine to arouse corticalcells. LeDoux notes that the arousal systems act in nonspe-cific ways throughout the forebrain to make cells moresensitive to incoming signals. In a sense, the nonspecificarousal interacts with the information processing of a par-ticular stimulus. The amygdala acts as a kind of centralprocessor and interacts with the prefrontal cortex (workingmemory and attention), hippocampus (long-term explicitmemory), and sensory cortex (perception and short-termstorage) to influence emotional responses. The amygdalanot only influences cortical areas of the brain but alsoreceives input from arousal networks (which themselvesalso influence the forebrain) and feedback from bodilyexpression of emotions. In addition, signals from theamygdalaaresent tomusclesand internalorgansandglands.

Most of the research to date into the role of arousal sys-tems and the amygdala in emotional behavior has beenlimited to a small number of emotions (e.g., fear). But it isbelieved that each emotional response is mediated by sepa-rate neural systems, although each may overlap or resem-ble each other in many respects. In sum, LeDoux (1996)and other brain researchers (e.g., Damasio 1994) construeemotions as biological functions of the nervous system(see also Zajonc 1998).

To our knowledge, there has been little work to eitherintegrate or reconcile cognitive theories of emotions withneural and biological theories. Much remains to be done inpsychological research before we can make definitivestatements about the precise role of arousal in emotionalexperience and behavior.

At least three studies have examined limited aspects ofarousal in marketing-related contexts. Sanbonmatsu andKardes (1988) found that arousal may govern attitude for-mation in persuasive message settings. Attitudes werebased on peripheral cues when respondents were highlyaroused but on argument strength when they were moder-ately aroused. It is unclear whether arousal functionedhere to reduce information-processing capacity of externalarguments or focused attention on internal reactions.

Bagozzi (1994) found that consistent with predictionsby knowledge-assembly theory (Hayes-Roth 1977),arousal transformed a two-dimensional, affective-cognitive representation of evaluations of giving bloodinto a one-dimensional, unitized representation. Likewise,arousal increased the association between attitudes andpositive beliefs about the consequences of giving bloodand decreased the association between attitudes and nega-tive beliefs. These predictions on the associations betweenattitudes and beliefs were explained by the implications ofspreading activation effects of arousal and copingresponses, wherein individuals attempt to avoid negativeand facilitate positive associations of beliefs with

attitudes. In another study, Bagozzi (1996) found that forattitudes toward giving blood, high arousal tends toenhance a halo effect from attitudes to positive beliefs andreduce the halo from attitudes to negative beliefs about theconsequences of giving blood.

Clearly, arousal is a fundamental aspect of behaviorrelated to emotions. We must acknowledge that appraisaltheories have not done a good, or at least complete, job ofincorporating arousal into their frameworks. In theirdefense, however, we should mention the following.Appraisal theorists recognize that the intensity of emo-tional experience consists of two components: arousal andself-control (e.g., Frijda 1994:120). Likewise, researchersaccept that “autonomic nervous system and other physio-logical processes” at least accompany subjectively feltemotions (e.g., Oatley 1992: 21) and that “[i]f the criterionof physiological activity was eliminated from the defini-tion, the concept of emotion would be left without one ofthe important response boundaries with which to distin-guish it from nonemotion” (Lazarus 1991:58-59). But isthere more to arousal in emotion than this?

Some psychologists and marketers have been quick todismiss Zajonc’s (1980) claim that “preferences need noinferences” (see Cohen and Areni 1991:215-216; Lazarus1982; Marcel 1983; Zajonc 1984). However, we believe itis important, at the present, to recognize that emotionalmeanings can be processed subconsciously, emotions canbe activated automatically, and responses to emotions(e.g., coping, action tendencies, actions) also can occurautomatically. We leave open the possibility that “emotionand cognition are best thought of as separate but interact-ing mental functions mediated by separate but interactingbrain systems” (LeDoux 1996:69; see also Oatley 1992,chap. 1). It appears that arousal is a key part of emotionalfunctions in the brain that underlies much of its automatic-ity. Cognitive appraisals and arousal need to be betterincorporated into our theories of emotion.

EMOTIONS AS MARKERS, MEDIATORS,AND MODERATORS OF CONSUMERRESPONSES

Based on content, most advertisements can be dividedinto two categories: (a) thinking ads, where focus is placedon either factual information (e.g., product attributes) orutilitarian consequences of product/service use (e.g., sav-ings in time or money) or (b) feeling ads, where concentra-tion is placed on the emotions one will experience throughuse or ownership of a product (see Puto and Wells’s [1984]similar distinction between informational and transforma-tional advertising). Rather than focusing on the stimulus,per se, it is more important to emphasize the processes andexperiences comprising a person’s response to ads, to


better understand the emotional meaning of ads (e.g., Fri-estad and Thorson 1986).

Paralleling the above differentiation between types ofads and the emotional-cognitive division in mentalprocesses mentioned earlier, Batra and Ray (1986) devel-oped a framework and coding scheme for classifyingaffective responses to ads, as a complement to cognitiveresponses. Specifically, Batra and Ray (1986) identifiedthree positive affective response categories: surgency-elation-vigor/activation (SEVA), deactivation, and socialaffection feelings. The SEVA feelings refer to upbeat,happy mood reactions (e.g., “the ad’s music was ‘catchy,’the ad was ‘fun to watch or breezy,’ or . . . alikable use ofhumor”); deactivation includes soothing, relaxing, quiet,or pleasing reactions; and social affection encompassesfeelings of warmth, tenderness, and caring (Batra and Ray1986:241). The three positive affective responses wereused along with six cognitive responses (i.e., support argu-ments, counterarguments, execution discounting, execu-tion bolstering, neutral distracters, and other reactions) ina study of the impact of television commercials on con-sumers. A total of 12 percent of reactions to ads were clas-sified as positive affect: SEVA (3.7%), deactivation(2.5%), and social affect (6.1%).

One use of emotional reactions in the above sensemight be as markers or indicators of the effectiveness ofadvertising copy, particularly with respect to the overallpersuasiveness of the ad, the appeal of spokespersons,evaluation of particular product claims, and appraisals ofother aspects of the execution (Wiles and Cornwell 1990).Also, the program surrounding an ad (e.g., happy versussad content) has been found to have main effects on one’sevaluation of an ad and recall (e.g., Goldberg and Gorn1987). Future research is needed to identify how programcontent and advertising appeals interact to influence con-sumer emotional responses.

A research question that has received quite a bit ofattention in recent years is how and to what extent emo-tional reactions to ads influence consumer decision mak-ing. Most often these reactions have been measured as atti-tudes (e.g., liking) toward the ad (e.g., Brown and Stayman1992; Mitchell and Olson 1981; Shimp 1981).

Attitude toward the ad (Aad) is thought to be a functionof feelings (and thoughts) about the ad itself (e.g., Batraand Ray 1986; MacKenzie, Lutz, and Belch 1986). In fact,Batra and Ray (1986) found that the three affectiveresponses discussed above significantly predicted Aadevenafter controlling for the effects of cognitive responses. Anumber of researchers have examined the conditionsunder which emotions influence Aad. For instance, underlow-involvement viewing conditions, feelings about thead have been found to be more important determinants ofAad than thoughts (e.g., how informative or useful the adis), but under higher involvement viewing, both feelingsand thoughts may be important (e.g., Miniard, Bhatla, and

Rose 1990). Presumably, high involvement promotes cog-nitive processing of the usefulness of the ad and itscontent.

A majority of research has addressed the effects of Aad,especially on attitudes toward the brand (Ab). Classic atti-tude theory maintains that Ab is a function of beliefs aboutbrand attributes or consequences of product use. Mitchelland Olson (1981) and Shimp (1981) were the first to findthat Aadprovided additional explanatory power for Ab overand above brand beliefs (see also Edell and Burke 1987).Batra and Ray (1986) found, however, that affectiveresponses toward the ad influenced Ab only indirectlythrough Aad.

Some research has addressed the conditions underwhich Aad influences Ab. For example, Brown and Stay-man (1992) revealed in their meta-analysis that the effectsof Aad on Ab are greater for novel than well-known brandsand for durable and other goods versus nondurables. Someevidence also exists showing that Aad influences Ab indi-rectly through its effect on beliefs about the brand (MacK-enzie et al. 1986). Finally, Stayman and Aaker (1988)showed that repetition governs the feelings to Aad relation-ship. Under levels of low versus high repetition, feelingshave a stronger effect on Aad. This may be a consequence ofgreater information processing under high versus lowrepetition.

In addition to the transfer of affect from ad emotions toAad, research shows that brand names and feelings towardads can become linked in memory. Stayman and Batra(1991) found that respondents exposed to an affective, asopposed to an argument, ad were able to retrieve brand atti-tudes faster, when primed with the brand name. In addi-tion, the positive retrieved affect had a stronger influenceon choice in low- than in high-involvement contexts. In asecond study, Stayman and Batra demonstrated that view-ers of an ad who were in a positive affective state morestrongly evoked the affect when given the brand name as asubsequent retrieval cue than viewers exposed to the adwhile not in a positive affect state.

Olney, Holbrook, and Batra (1991) investigated a hier-archical model explaining advertising viewing time. Theeffects of ad emotions (i.e., pleasure and arousal) weremediated by Aad and reactions to ad content. Interestingly,arousal had both indirect and direct effects on viewingtime, even after controlling for ad content and Aad. Arousalwas measured by self-reports.

Emotions have been found to serve as moderators intheir impact on Ab. Batra and Stayman (1990), in one of thefew studies to examine mood and print ads, found thatpositive moods enhance Ab through their interaction withtwo cognitive processes: “(1) a bias against the generationof negative thoughts (such as evoked by weak arguments),leading to a more favorable evaluation of message argu-ments, and (2) a reduction in total cognitive elaboration,making processing more heuristic than systematic” (pp.


212-213). An interesting finding was that positive moodsseem to reduce counterargumentation when weak argu-ments are used in ads (see also Worth and Mackie 1987).Batra and Stephens (1994) also investigated the moderat-ing effects of mood on Ab. Specifically, they showed thatmood and motivation (the latter conceived as degree ofrelevance of the product category for consumers) inter-acted to govern Ab when consumers watched televisionads. The greatest impact on Ab occurred when positivemoods ensued under conditions of low motivation. Therationale is that positive moods and low motivation sup-press counterargumentation (and therefore lead to morefavorable Ab) in comparison with high-motivationconditions.

Following the elaboration likelihood model (ELM)(e.g., Petty and Cacioppo 1986), we might expect emo-tions to have one or both of two effects. Emotions mightoperate centrally to influence cognitive processes (e.g.,with regard to argumentation), or emotions might functionperipherally (perhaps through associative or affect transfermechanisms). Early predictions under the ELM took aneither-or perspective and stressed that when the processingof information in a communication is low (e.g., due to lowmotivation, distraction, low need for cognition, weakarguments), emotional content in the communication(e.g., an attractive spokesperson) is processed directly andtransfers to, or influences, attitude toward the product ormessage. When issue-relevant thinking is high, attitudechange is thought to be a function of the balance of prosand cons in the communication, a largely rational process,and emotion may not be a factor.

Over time, the role of emotion in persuasive communi-cation has been found to be more complex than the simplecentral versus peripheral processing alternatives specifiedin the ELM (e.g., Wegener and Petty 1996). For example,when the likelihood of information processing is moder-ate, emotions have been found to affect the extent to whicharguments in a communication become elaborated. Posi-tive (versus neutral) moods tend to lead to less processingof arguments (e.g., Bless, Bohner, Schwarz, and Strack1990; see also discussion below on the effects of emotionson cognitive processes). By contrast, when people processthe arguments in a message closely, mood might biasinformation processing or even function as an argumentitself (Forgas 1995). Another explanation for mood effectsis that sad or neutral (versus happy) moods lead to moreeffortful processing, which is believed to be done sponta-neously (e.g., Bohner, Chaiken, and Hunyadi 1994). Notethat this prediction seems to conflict with the observationsmade by Schaller and Cialdini (1990), who analyzedmostly nonpersuasive communication studies. Some clari-fication for the discrepancy in interpretations can be seenin a study by Wegener, Petty, and Smith (1995). Wegeneret al. (1995) discovered that happy (versus sad) moodslead to the processing of more arguments in a message

when a “proattitudinal/uplifting” position was taken, buthappy (versus sad) moods lead to less processing of argu-ments when a “counterattitudinal/depression” positionwas taken. Wegener et al. (1995) explain the findings, inpart, by suggesting that when in a happy mood, people tryto maintain their mood and thus process less of the coun-terattitudinal/depression content.

In addition, interactions of emotions sometimes occurwith other variables, such as motivational or ability fac-tors. For instance, Wegener, Petty, and Klein (1994) foundthat for people high in need for cognition, messagesframed positively (i.e., good things will happen if peopleadopt the advocacy) were more persuasive when the audi-ence was happy (versus sad), but messages framed nega-tively (i.e., bad things will happen if people do not adoptthe advocacy) were more persuasive when the audiencewas sad (versus happy). Wegener et al. (1994) explainedthese findings by claiming that good outcomes seem betterand more likely to occur while in a good (versus sad)mood, and bad outcomes seem worse and more likely tooccur while in a sad (versus happy) mood.

Finally, even when the likelihood of information pro-cessing is low (due, e.g., to low need for cognition or lowrelevance of a product for a consumer), the mood of theaudience can have a direct effect on Ab, but when the likeli-hood of information processing is high, the mood of theaudience affected message-generated thoughts consistentwith the mood (Petty, Schumann, Richman, andStrathman 1993). Mood was induced by a television pro-gram or music.

THE EFFECTS OF EMOTION 1:INFLUENCES ON COGNITIVEPROCESSES

A person’s emotional state can influence variousaspects of information processing including encoding andretrieval of information, different strategies used toprocess information, evaluations and judgments, and crea-tive thinking. In this section we examine the influence ofaffective states on various aspects of cognitive processes.

Emotion/Mood Effects on Memory

The influence of mood states on memory can bebroadly classified into three categories: retrieval effects,encoding effects, and state-dependent learning effects. Inthe following section, we review some of the studies thathave investigated these effects and discuss the mechanismby which affect influences memory.

Retrieval effects. Affect has been shown to influenceretrieval of information, whereby persons in a positivemood state at the time of retrieval have been found to show


superior recall of positive material learned during encod-ing, relative to neutral or negative material (Isen, Shalker,Clark, and Karp 1978; Laird, Wagener, Halal, and Szegda1982; Nasby and Yando 1982; Teasdale and Russell 1983).For example, Isen et al. (1978) had respondents study posi-tive, negative, and neutral words. Either positive, neutral,or negative mood states were induced in these respon-dents. Respondents in the positive-mood condition re-trieved more positive words compared with neutral ornegative words. Isen et al. (1978) suggest that thinkingabout mood-incongruent material involves shifting one’sfocus, which is cognitively taxing, and therefore peopleare more likely to focus on mood-congruent material. An-other mechanism proposed to explain the retrieval effectsof positive affect suggests that positive mood at the time ofretrieval functions as a cue that primes the positive mate-rial in memory, making these material more accessible(Isen 1989; Isen et al. 1978; also see Tulving and Pearl-stone 1966 for a discussion on the effects of priming on ac-cessibility). The easier accessibility of positive materialmay then influence other cognitive processes such asevaluations and decision making, and also subsequent be-haviors. While retrieval effects have been replicated bymany researchers using different mood induction and test-ing techniques, a few prominent studies have failed to de-tect retrieval effects (Bower, Monteiro, and Gilligan 1978;Bower, Gilligan, and Monteiro 1981), leading Isen (1984)to speculate that this failure may have been caused by thespecific material and induction methods (such as hypno-sis) used in these studies.

Encoding effects. Mood states have also been shown toexhibit encoding effects whereby the affective state at thetime of learning is associated with superior memory forsimilarly valenced material (Bower and Cohen 1982; For-gas and Bower 1987). Nasby and Yando (1982) found thatpositive mood at the time of learning led to an improved re-call of positive material at a later point in time regardless ofthe mood state at the time of recall. Bower et al. (1981)found evidence for the encoding effect of both positive andnegative affect. Respondents were made to feel happy orsad and then read descriptions of various psychiatric inter-views. Happy respondents learned many more happy factsthan sad facts, while sad respondents learned many moresad facts than happy facts.

How can the encoding effects of mood be explained?Bower and colleagues (e.g., Bower and Cohen 1982) havesuggested that mood-congruent material is likely to bemore semantically elaborated relative to mood-incongruent material. Forgas and Bower (1987) found thatin impression formation situations, sad individuals spentmore time examining negative rather than positive infor-mation, and they subsequently recalled the negative infor-mation better. Conversely, happy individuals spent moretime on the positive information and recalled that better at

a later time. The greater levels of associations evoked bymood-congruent material may have caused a more exten-sive elaboration, which, in turn, requires more time. How-ever, Isen et al. (1978) and Srull (1983) failed to find anyencoding effects of affect.

State-dependent learning effects. A third memory ef-fect of mood is the state-dependent learning effect of af-fect, where any material regardless of its affective valencelearned under a particular mood state is recalled betterwhen the person is again in that affective state (Bartlett,Burleson, and Santrock 1982; Bartlett and Santrock 1979;Bower, Monteiro, and Gilligan 1978; Bower, Gilligan, andMonteiro 1981). Bower et al. (1978) had respondents learntwo sets of words, one while they were in a positive affec-tive state and the other in a negative mood. When respon-dents who learned the two lists in different moods recalledthe words in the wrong mood (e.g., when words learned ina positive mood were recalled while respondents were in anegative mood), they experienced interference and the av-erage recall rate was less than 50 percent. When respon-dents who learned two lists in different moods recalled thewords in the correct mood, the average recall rate wasmore than 70 percent. Control respondents who learnedand recalled both lists while in the same mood showed anaverage recall rate between 50 and 60 percent. Bower andCohen (1982) suggest that the respondents’ mood at thetime of learning becomes associated with the learned ma-terial and that these associations facilitate the recall oflearned material when the mood state at recall matches themood state at encoding (also see Bower 1981 for a descrip-tion of his semantic-network theory).

Evidence for mood state-dependent learning has beenambiguous. Many studies have failed to find any state-dependent effects of mood state (Bower and Mayer 1985;Isen et al. 1978; Laird et al. 1982; Nasby and Yando 1982).Eich and Birnbaum (1982) and Isen (1984, 1989) havesuggested that when the material to be learned has seman-tic meaning, the stimulus will be encoded according to thismeaning, and the influence of the mood state in the encod-ing and subsequent retrieval processes will be minimal.However, when the stimulus lacks meaning, contextualcues such as affective states at the time of learning may bemore strongly encoded with the learned material. At thetime of retrieval, these memory items, which have fewsemantic associations, are more primed by the matchingaffective state at recall.

Asymmetric effects of positive and negative moods.While positive affective states have been shown to havesignificant influences on recall, negative affect has some-times been found to have either no effect or a much smallereffect on the recall of negative material from memory.Asymmetric effects of positive and negative moods havebeen found for retrieval effects (Isen et al. 1978; Nasby


and Yando 1982; Teasdale and Fogarty 1979), encodingeffects (Nasby and Yando 1982), and state-dependentlearning effects (Bartlett and Santrock 1979; Bartlett et al.1982). Isen (1984) speculated that positive affect is struc-tured in a broad and extensive manner (i.e., highly inter-connected with other memories), while negative affect ismore narrowly and less well connected with other mate-rial, and that specific negative affective states such as an-ger and sadness may be organized separately in memory.This, in turn, would make it difficult for any given negativemood to act as an effective retrieval cue. It is not readily ap-parent why negative affect would be less well connected inmemory and positive affect more widely interconnected asIsen suggests, considering that negative affect usually sig-nals problematic environmental conditions that may re-quire problem solving (Schwarz and Clore 1983; Wegeneret al. 1995). One of the side effects of this hypothesized in-terconnectedness of positive memories is the greater crea-tiveness and cognitive flexibility demonstrated by peoplein positive moods, a topic we review briefly later in thisarticle.

Mood maintenance and repair have also been put for-ward as explanations for the asymmetric effects of positiveand negative moods. Isen (1984) has suggested that happyindividuals attempt to prolong their positive affective stateby focusing on the positive aspects of their stimulus (moodmaintenance), while individuals in a negative mood try toimprove their situation by not focusing on negative memo-ries (mood repair). Isen (1989) noted that in some of thestudies showing symmetrical effects of positive and nega-tive mood (Bower et al. 1978, 1981), respondents wereinstructed to maintain their induced moods, and this mayhave discouraged them from engaging in mood repairstrategies. However, the mood repair explanation is notwithout problems. If sad respondents engage in moodrepair, why is there no evidence for mood-incongruentrecall effects? After all, an effective strategy to improve adepressed mood state is to engage in pleasant thoughts andmemories. Yet, evidence seems to show that negativemood inhibits the recall of positive memories (e.g., Isen etal. 1978). The competing explanations of memory struc-ture differences and mood maintenance/repair have alsobeen used to explain differences in information-processing strategies of happy and sad individuals, a topicwe discuss later in this article.

Mood, Categorization, and Creativity

Various studies investigating the influence of mood oncategorization have found that people in positive moodstates, compared with those in neutral or negative moodstates, tend to be better at integrating information, findingrelationships among stimuli, and at finding creative

solutions (Isen and Daubman 1984; Isen, Daubman, andNowicki, 1987; Isen, Johnson, Mertz, and Robinson 1985;Isen, Niedenthal, and Cantor 1992; Murray, Sujan, Hirt,and Sujan 1990). For example, respondents in positivemood conditions tended to group a wider range of neutralstimuli together (Isen and Daubman 1984). They alsorated words such ascane, ring, andpurseas being betterexemplars of the category “clothing,” than did the neutral-mood respondents. Murray et al. (1990) found thatpositive-mood respondents, compared with respondentsin other mood states, formed broader categories whenfocusing on similarities among exemplars and narrowercategories when focusing on differences, prompting themto suggest that what positive mood promotes is not broadercategorization but rather cognitive flexibility. Isen et al.(1985) found that positive-mood respondents tended togive more unusual responses to neutral words in wordassociation tests. For example, in response to the wordhouse, positive-mood respondents were more likely thanwere neutral-mood respondents to mention unusual firstassociates such assecurity, residence, andapartment, sug-gesting that positive-mood states may influence cognitiveorganization, resulting in more flexible interpretation ofrelationships among stimuli. This cognitive flexibility alsoresults in enhanced creativity. Isen et al. (1987) found thatrespondents in positive-mood states outperformed those inneutral- and negative-mood states on tests requiring crea-tive solutions. These researchers suggest that positive-mood respondents were better at creative problem solvingsince such tasks required the ability to see relatednessamong seemingly unrelated stimuli, and as we have seenearlier, positive affect results in more flexible cognitiveorganization.

Mood Effects on Evaluation

One of the best-recognized and most robust effects ofmood is its influence on evaluation. Individuals inpositive-mood states have been shown to evaluate stimulimore positively than individuals in neutral- or negative-mood states, whether the stimuli being studied are otherpeople (Clore and Byrne 1974; Forgas and Bower 1987),consumer goods (Isen et al. 1978; Srull 1983), life satis-faction (Schwarz and Clore 1983), or past life events(Clark and Teasdale 1982). The reliability of mood effectson evaluation is perhaps best illustrated by the fact thatevaluations of the pleasantness of neutral/ambiguousstimuli are used as checks for mood manipulation (e.g.,Isen et al. 1985, 1987). Isen et al. (1978) found that respon-dents in whom positive mood was induced were morelikely to rate their cars and televisions more favorablycompared with respondents in neutral-mood states. In thecase of memory-based evaluations, if the recalled infor-mation is biased by the mood (as discussed in the


following section), then evaluations that follow will bebiased too. On-line processing of evaluations can also beinfluenced by mood states through the retrieval of infor-mation congruent with the mood (Clore, Schwarz, andConway 1994).

A competing explanation based on the feelings-as-information model (see Schwarz 1990; Schwarz and Bless1991; Schwarz and Clore 1983) suggests that individualsmay assume that their mood states are affective reactionsto the object being evaluated and thus base their evalua-tions on their affective states. For example, a happy indi-vidual when asked to evaluate a painting may ask the ques-tion, “How do I feel?” and infer that his or her positivemood is a reaction to the painting and therefore come to theconclusion that he or she likes the painting. The feelings-as-information hypothesis suggests that when individualsattribute their mood state to something else other than theobject being evaluated, the effect of mood on evaluationshould disappear. Schwarz and Clore (1983) found sup-port for this hypothesis when they showed that althoughpeople called on sunny days reported more life satisfactionthan people called on cloudy days, the differences disap-peared when the interviewer casually mentioned theweather to the individuals. Presumably, the casual men-tion of the weather made people attribute their mood to theweather, and hence the mood lost any diagnostic value inevaluating life satisfaction. In other words, people may usetheir moods as the basis for forming evaluations of objectsunless the diagnostic value of the mood is discounted.Clore et al. (1994) compared the two explanations formood effects on evaluation and suggested that individualsmay use feelings as information when the evaluation taskis affective in nature, when other information is lacking,when the information is complex, or when there are timeconstraints.

Although mood effects on evaluation have been repli-cated often, a few prominent studies have shown that moodstates will not influence evaluation when the object beingevaluated is highly familiar and for which past evaluationsexist in memory (e.g., Salovey andBirnbaum 1989; Srull1983, 1984). Srull (1984, Experiment 3) found that theevaluations of a car made by novices, but not by experts,were influenced by mood state. Novices, by definition, areunfamiliar with the product category and are more likely toengage in on-line evaluations, which are more susceptibleto mood influences. Experts, on the other hand, have priorevaluations available in memory and so do not engage inon-line evaluations and are less likely to be influenced bymood. Similar findings have been obtained by Salovey andBirnbaum (1989, Experiment 3) and Schwarz, Strack,Kommer, and Wagner (1987). We discuss this and otherrelated issues in greater detail later in this article.

Mood Effects on Information Processing

It has been suggested that the mood maintenance strat-egy used by people in positive moods may also cause themto avoid investing cognitive effort in tasks unless doing sopromises to maintain or enhance their positive mood (e.g.,Isen 1987; Wegener et al. 1995). Consequently, people inpositive-mood states may not be motivated to engage insystematic processing of information and may use heuris-tic processing instead. Positive affect usually denotes abenign environment that does not require any action. Onthe other hand, negative affective states act as informationsignaling that the environment poses a problem and maymotivate people to engage in systematic processing, whichis usually better suited to handling threatening situations(Schwarz 1990; Schwarz and Clore 1983). Various stud-ies have found evidence suggesting the use of heuristicprocessing by people in positive moods and systematicprocessing by people in negative moods (e.g., Bless et al.1990, 1996; Mackie and Worth 1989). Bless et al. (1990)presented happy and sad individuals with either strong orweak counter-attitudinal arguments. Sad individuals wereinfluenced only by strong arguments, while happy indi-viduals were equally influenced by strong and weak argu-ments. These effects have been consistently replicated andhave been interpreted as providing evidence for reducedsystematic processing by individuals in positive affectivestates. The two major mechanisms that have been used toexplain mood effects on information-processing strategiesare the same as those used to explain the asymmetriceffects of mood effects on memory, namely, the highlyinterconnected nature of positive memories and moodmaintenance.

In discussing the asymmetric effects of positive andnegative affect on memory, we had briefly discussed Isen’s(1984) contention that positive concepts are more highlyinterconnected in memory relative to negative ideas. Isen(1987) and Mackie and Worth (1989) have argued thatsince positive memories are highly interconnected, posi-tive mood will prime and activate some related and manyunrelated positive memories, thus leading to cognitivecapacity constraints. Because individuals do not have thecognitive resources to engage in systematic processing,they resort to the less demanding heuristic processing.

A second explanation for mood effects on cognitiveprocessing is based on the concept of mood maintenance.Isen (1987) has suggested that individuals in a positiveaffective state are motivated to maintain their mood andmay avoid cognitive activity that could interfere with theirpositive mood. Thus, individuals in a happy mood areunlikely to engage in systematic processing of information(also see Bohner, Crow, Erb, and Schwarz 1992).

Another motivation-based explanation for the reducedprocessing under positive mood has been offered bySchwarz and colleagues (e.g., Schwarz and Clore 1983)using their feelings-as-information hypothesis. Negative


affective states inform people that they may be facing aproblem and this may provoke systematic processing ofinformation that is better suited to problem solving. Posi-tive mood, on the other hand, informs the individual thatthe environment is benign and thus he or she may not bemotivated to engage in effortful cognitive processing. Afourth hypothesis suggests that the motivation to simplifyprocessing is not what causes the reliance on heuristicprocessing; rather, it is the increased use of heuristic pro-cessing by these happy-mood individuals that results insimplified processing (e.g., Bless et al. 1996).

Although the increased reliance on heuristics by indi-viduals in positive moods has been replicated many times,there have been studies that have shown that positivemoods could lead to both increased or diminished levels ofcognitive processing (e.g., Martin, Ward, Achee, and Wyer1993). Wegener et al. (1995) have presented empiricalsupport for their hypothesis that positive moods could leadto either lesser or greater levels of information processing,depending on whether happy individuals believe that sys-tematic processing will lead to the maintenance or destruc-tion of their positive mood. People in a positive-moodstate, who believe that systematic processing of a messagewould help maintain their mood, may engage in moredetailed processing. However, if they view systematicprocessing of that message as a threat to their positive-mood state, they would avoid elaborate processing. Thishedonic contingency mood management is especiallylikely to be seen in positive-mood individuals who havethe most to lose by incorrectly engaging in systematicprocessing. Sad individuals, by contrast, have less to lose,since processing messages might make them feel better(Wegener and Petty 1994). Wegener et al. (1995) suggestthat most of the previous research in which positive moodshave been shown to lead to reduced message processinghave involved counterattitudinal or otherwise unpleasantcognitions (e.g., Mackie and Worth 1989). Since the sys-tematic processing of counterattitudinal or unpleasantmessages is unlikely to help maintain a positive mood,individuals choose not to use systematic processing.

Forgas (1994, 1995), building on Fiedler’s (1990) dual-force model, has recently presented theaffect infusionmodel, which suggests that the various explanations formood effects on memory and information processing(such as mood maintenance, feelings as information,affect priming) are not necessarily competing models butcould rather be complementary explanations. He suggeststhat affect infusion into judgmental processes takes placewhen the judgments require a high degree of constructive(on-line) processing. Affect does not influence those judg-ments that call for the retrieval of preexisting evaluationsor for information processing that is highly specific andnot requiring constructive processing. Forgas (1994,1995) suggests thatdirect-access processing(a low-affectinfusion strategy) is used when the evaluative target is

familiar and when there are no strong cognitive, affective,or situational factors that call for systematic processing.Accordingly, individuals who evaluate very familiarobjects would be using direct-access processing, andhence mood effects would not be found in such cases (cf.Srull 1984). Another low-affect infusion strategy ismoti-vated processing, where the information search and theevaluative outcome are guided by prior motivational goalssuch as mood repair (see Forgas 1995 for a detailed discus-sion). When the evaluative target is simple, the personalrelevance is low, the individual has limited cognitivecapacity, or the accuracy requirements are not high, theindividual may resort toheuristic processing, a high-affectinfusion strategy. Under these circumstances, evaluationsmay be based on the existing mood, as in the feelings-as-information hypothesis (cf. Schwarz and Bless 1991).When the judgmental situation requires the individual tolearn and process novel information, and when he or shehas adequate cognitive capacity and motivation to processthe information,substantive processingis predicted totake place. According to Forgas (1995), this default strat-egy, which is used if other less effortful strategies areinadequate, involves constructive processing and is onethat is susceptible to affect infusion (cf. Srull 1984).

THE EFFECTS OF EMOTIONS 2:INFLUENCE ON VOLITIONS,GOAL-DIRECTED BEHAVIOR,AND DECISIONS TO HELP

Sometimes emotions spur one onto action; at othertimes emotions inhibit or constrain action. But onlyrecently have researchers devoted much attention to study-ing how this occurs.

Bagozzi, Baumgartner, and Pieters (1998) investigatedthe role of anticipatory emotions in goal striving. In apanel study of consumers’ efforts to lose or maintain theirbody weight, people first reacted to the possibility ofachieving or not achieving their goals with well-definedpositive and negative anticipatory emotions. The anticipa-tory emotions then energized volitions in the form ofintentions, plans, and the decision to expend energy in theservice of goal striving. Volitions, in turn, were shown toinfluence goal-directed behaviors related to exercisingand dieting. Next, the intensity of execution of thesebehaviors contributed to degree of goal attainment. Goal-outcome emotions subsequently followed.

Anticipatory emotions functioned in the followingmanner. Respondents indicated which emotions theyanticipated they would experience if they were to achievetheir goal (i.e., excitement, delight, happiness, gladness,satisfaction, pride, and self-assurance) and which emo-tions they anticipated they would experience if they wereto fail to achieve their goal (i.e., anger, frustration, guilt,


shame, sadness, disappointment, depression, worry,uncomfortableness, and fear). The key processes are simi-lar to forward-looking counterfactual thinking processes(e.g., Roese and Olson 1995) or what Gleicher et al. (1995)term “prefactuals.” The decision maker begins by “imag-ining the possible” (i.e., identifying and evaluating theconsequences occurring if one were to achieve one’s goalor not). The alternative consequences to imagined goalsuccess and goal failure serve as input to appraisalprocesses and felt emotional experiences. People are thenmotivated to choose actions promoting the positive affectand avoiding the negative affect associated with goalattainment and goal failure, respectively.

Four other studies support the above role for prefactualsin decision making. Boninger, Gleicher, Hetts, Armor, andMoore (1994) discovered that prefactuals with regard tothe purchase of insurance in a laboratory game influencedsubsequent action to take insurance. Gleicher et al. (1995)found that prefactuals concerning condom use affectedpositive attitudes; however, intentions to use condomswere not influenced by prefactuals, although the differ-ence in means between experimental and control groupswas in the predicted direction. Perugini and Bagozzi(1998) also showed that anticipated emotions were instru-mental in influencing desires, volitions, and behavior in astudy of exercising and an investigation of studying behav-iors (see also Parker, Manstead, and Stradling 1995; Rich-ard, van der Pligt, and de Vries 1995). Finally, Brown,Cron, and Slocum (1997) replicated the model proposedby Bagozzi, Baumgartner, and Pieters (1998) in a study ofgoal pursuit by salespeople.

Although emotions often function in broad categoriesof positive and negative affect, specific subcategories mayhave special relevance in certain contexts. Verbeke andBagozzi (1998) studied the behavior of 458 salespeople inEurope who sold financial services and found that twonegative emotions were particularly important: shame andembarrassment. These self-conscious emotions encom-pass personal standards with regard to acceptablethoughts, feelings, and actions. Shame occurs when onefails to live up to an ideal and another person whoseapproval is important to oneself judges this failure criti-cally. Embarrassment is generated either when the selfdoes something inappropriate in public or the other withwhom one interacts does something inappropriate, but incontrast to shame, the threat under embarrassment is lessfundamental and involves relatively trivial and evenhumorous behaviors (e.g., passing gas audibly while in asales negotiation). With embarrassment, the threat is to the“presented self,” and one feels foolish and awkward; withshame, the threat is to the “core self,” and one feels quitepained, perhaps even humiliated or devastated. Verbekeand Bagozzi (1998) found that salespersons’propensity toexperience shame and embarrassment led to particularprotective or coping responses: the greater the shame and

embarrassment, the greater the impulse to hide, to with-draw, and to avoid contact with customers. These copingresponses, in turn, negatively influenced salesperformance.

The study by Verbeke and Bagozzi (1998) illustratesdysfunctional effects of negative emotions (see also Ver-beke and Bagozzi 1999). Yet, negative emotions can havepro-social consequences in certain instances. Bagozzi andMoore (1994) investigated the role of negative emotionsand empathy on decisions to give help to abused children.In Study 1, the effects of a negative emotional-appeal adwere compared with those of a rational-appeal ad; inStudy 2, the effects of three levels of intensity of negative-emotion appeals were examined. The negative emotionalads featured a young boy in his home running away fromhis father in terror. Both studies demonstrated that fournegative emotions (anger, sadness, fear, tension) and fouraspects of empathy (perspective taking, compassion/pity,protection motivation, fantasy elaboration) mediated theeffects of viewing negative emotional ads on decisions tohelp. Stronger felt negative emotions in the audience led togreater feelings of empathy, and this, in turn, enhanced thedecision to help victims of child abuse.

A final area where emotions play a key role in promot-ing action is in goal setting. We can think of goal setting interms of a hierarchy of three classes of goals (Bagozzi andDholakia 1998; Pieters, Baumgartner, and Allen 1995). Afocal or basic-level goal answers the question, “What is itthat I strive for?” For example, a consumer may have a goalto lose body weight. The means for achieving this goal canbe interpreted as subordinate goals and answer the ques-tion, “Howcan I achieve that for which I strive?” Varioustypes of exercising behaviors and dieting activities areexamples of subordinate goals for the focal goal of losingweight. Superordinate goals constitute reasons for pursu-ing a focal goal and answer the question, “Whydo I want toachieve that for which I strive?” Bagozzi and Edwards(1998) investigated the hierarchical structure for super-ordinate goals governing body weight maintenance andfound that happiness was an important goal toward whichthe focal goal and other more concrete superordinate goalspointed (other superordinate goals included, for example,social acceptance, self-esteem, look good, feel good, andfit into clothes). Happiness, or more generally personalwelfare, seems to be a central motive spurring one ontoaction. We turn now to the role of happiness and otheremotions in customer satisfaction.

EMOTIONS AND CUSTOMERSATISFACTION

Early research viewed customer satisfaction as a keyoutcome of product/service purchase, whereby a


comparison is made between expectations of performanceand actual performance, and satisfaction arises whenactual performance is greater than or equal to expectedperformance, and dissatisfaction occurs otherwise (e.g.,Oliver 1980; Yi 1990). Westbrook (1987) explored theinfluence of positive affect and negative affect on satisfac-tion, along with expectation beliefs and disconfirmationbeliefs. Satisfaction with automobiles was a function ofpositive affect, negative affect, expectations of receivingbenefits and liabilities, and disconfirmation beliefs. Satis-faction with cable television was determined by positiveaffect, negative affect, and disconfirmation beliefs. West-brook and Oliver (1991) found that three emotionalresponses were important antecedents of satisfaction ofnewly purchased automobiles: pleasant surprise, interest,and hostility.

Oliver (1993) expanded the determinants of satisfac-tion to include positive affect (interest and joy) and nega-tive affect (anger, disgust, contempt, shame, guilt, fear,sadness), as well as disconfirmation beliefs. In addition,attribute satisfactions and attribute dissatisfactions werehypothesized as direct determinants of satisfaction, as wellas indirect (through positive and negative affect) determi-nants. Satisfaction with automobiles was found to be afunction of attribute satisfaction, attribute dissatisfaction,positive affect, negative affect, and disconfirmation. Satis-faction with a required university course was also found tobe determined by these antecedents, except attributedissatisfaction.

The investigation of the impact of emotions on post-purchase reactions is an important development in market-ing. However, it is unclear whether satisfaction is phenome-nologically distinct from many other positive emotions.Satisfaction is neither a basic emotion nor a centralemotional category in leading theories of emotions (e.g.,Frijda 1986; Lazarus 1991; Oatley 1992; Roseman 1991;Smith and Ellsworth 1985). Furthermore, Shaver,Schwartz, Kirson, and O’Connor (1987) found that satis-faction shares much common variance with such positiveemotions as happiness, joy, gladness, elation, delight, andenjoyment, among others. Likewise, Nyer (1997b) discov-ered that measures of joy and satisfaction loaded on onefactor.

Although we leave open the possibility that measures ofsatisfaction can achieve discriminant validity from mea-sures of joy, happiness, and other positive emotions, webelieve that this will be very difficult to produce in prac-tice. Also, no theory exists for specifying the conditionsunder which satisfaction exists uniquely from many otherpositive emotions. We suspect that previous studies find-ing discriminant validity for measures of satisfaction canbe explained by the way items were presented on the ques-tionnaire (e.g., separation of measures of satisfaction from

measures of other positive emotions) or the lack of inclu-sion of a sufficient number of positive emotions. No studyto date has performed either a rigorous experimental orsurvey (e.g., multitrait-multimethod) examination of con-struct validity of measures of satisfaction, joy, and relatedpositive emotions.

The centrality of satisfaction in marketing studies isperhaps more due to being the first emotion to receivescrutiny in postpurchase behavior research than to consti-tuting a unique, fundamental construct in and of itself.Indeed, it is likely that—depending on the situation, prod-uct, and person—other positive and negative emotions aremore important outcomes of purchase. Under certain con-ditions, frustration, anger, disappointment, alienation, dis-gust, anxiety, alarm, guilt, shame, joy, happiness, hope,pride, jubilation, excitement, relief, amusement, andpleasure, among many other negative and positive emo-tions, might be more valid reactions consumers have topurchases. By the same token, the implications of emo-tional reactions in purchase situations on complaintbehaviors, word-of-mouth communication, repurchase,and related actions may differ for various positive andnegative emotions and be of more relevance than reactionsto satisfaction or dissatisfaction, per se.

Research by Nyer (1997a, 1997b) addresses appraisaltheories and their role in postconsumption responses.Nyer found that such postconsumption responses as repur-chase intentions, word-of-mouth intentions, and otherreactions are predicted best by using measures of satisfac-tion plus measures of other emotions. Other studies of noteinvestigating the role of specific emotions in customer sat-isfaction include those done by Dubé, Bélanger, andTrudeau (1996); Folkes, Koletsky, and Graham (1987);and Taylor (1994).

We are uncertain whether a single, summary emotionalresponse such as satisfaction is feasible or even desirable.But if one is to be discovered, it may lie more in emotionsmore closely connected to human welfare or emotionalwell-being (e.g., Diener and Larsen 1993). In this regard,for example, a case could be made for happiness as a fun-damental emotion related to the purchase of goods and ser-vices, in particular, and emotional well-being, in general(Bagozzi forthcoming).

SOCIAL BASES OF EMOTIONS

Although people can experience emotions privately,such as in response to physical danger, a case can be madethat emotions are most often interpersonal or group-basedresponses. Unfortunately, the vast majority of researchinto emotional behavior has had an individualistic slant toit (e.g., Parkinson 1995). Marketing relationships seem tobe contexts where more social conceptualizations of


emotions would be worth pursuing. For example, Ruth,Otnes, and Brunel (1998) review studies where discreteemotions are central in gift exchanges, and they show howappraisals lead to emotions and how emotions relate tointerpersonal relationships and disposition of gifts. Arelated area in need of study is the management of emo-tions by organizations and by the self (e.g., Bagozzi 1992;Hochschild 1983; Locke 1996). Finally, cultural aspects ofemotions deserve further inquiry (e.g., Markus and Kita-yama 1994).

CONCLUSIONS

Emotions are ubiquitous throughout marketing. Theyinfluence information processing, mediate responses topersuasive appeals, measure the effects of marketing stim-uli, initiate goal setting, enact goal-directed behaviors, andserve as ends and measures of consumer welfare. Yet, weare only beginning to understand the role of emotions inmarketing.

The following areas constitute opportunities for futureresearch:

Area No. 1. Exactly how are appraisals conducted andhow do they lead to emotional reactions in consumers? Inwhat sense are appraisals conscious and purposive versusautomatic? What is the role of the amygdala, hippocam-pus, and other neural systems in appraisals? What are theessential elements of cognitive appraisals in emotionalbehavior and how can they be influenced by marketingstimuli?

Area No. 2. Related to No. 1, but also more generallythroughout the processes involved in emotional responses,what role does arousal play? What is physiological arousaland how does it relate to appraisals, coping responses,action tendencies, and behavior? What do self-reports ofarousal really indicate? Is arousal an essential componentof emotions? If so, how can marketers develop theories ofemotions, better incorporating arousal, and how shouldarousal be measured?

Area No. 3. How do emotions affect information pro-cessing in consumer decision making? In what ways doemotions influence the encoding, storage, and retrieval ofinformation? What contribution do emotions make to con-sumer decision making and choice?

Area No. 4. Are emotional reactions universal? In whatways do emotions differ across cultures? What role dosocialization, ethnicity, and culture play in the representa-tion, experience, and effects of emotions?

Area No. 5. Under what conditions do emotions func-tion in discrete categories and what are the implications ofdiscrete emotions for consumer behavior? When do emo-tions function in clusters or as homogeneous groups, suchas positive and negative categories?

Area No. 6. How do emotions function to influence thebehavior of salespeople and managers? When are theyfunctional or dysfunctional?

Area No. 7. How do consumers and managers controltheir emotions to advantage? What role do emotions playin self-regulation?

Area No. 8. How should emotions be measured in mar-keting? When are self-reports appropriate and inappropri-ate? How can physiological measures be incorporated?What are the consequences of treating emotions as unipo-lar versus bipolar responses?

Area No. 9. What are the distinctions between and rela-tionships among emotions, affect, feelings, evaluations,moods, and attitudes?

Area No. 10. How are distinct emotions related to eachother? Under what conditions, for example, does frustra-tion lead to dissatisfaction? Shame lead to anger? Or lovelead to happiness?

Area No. 11. An area neglected by marketers, but at theheart of the discipline, is the role of emotions in marketingexchanges and relationships. How do emotions initiate,maintain, or sever marketing relationships? Can emotionsand marketing relationships be studied more dynamicallyas they evolve in real time and in context? What are theimplications of treating emotions in marketing as socialphenomena as opposed to strictly intrapsychicphenomena?

Area No. 12. Finally, is customer satisfaction a uniquephenomenon or is it a subcategory of positive emotions?Likewise, is dissatisfaction a unique phenomenon, thepolar opposite of satisfaction, or a subcategory of negativeemotions? How do satisfaction/dissatisfaction and otheremotions relate to customer loyalty, complaint behavior,and word-of-mouth behavior? What role do emotions playin the diffusion of innovations?

ACKNOWLEDGMENTS

The authors would like to thank Professor Julie Ruthfor her comments on this article.

REFERENCES

Aaker, David A., Douglas M. Stayman, and Michael R. Hagerty. 1986.“Warmth in Advertising: Measurement, Impact and Sequence Ef-fects.”Journal of Consumer Research12 (March): 365-381.

, , and Richard Vezina. 1988. “Identifying Feelings Elic-ited by Advertising.”Psychology & Marketing5 (Spring): 1-16.

Allen, Chris T. and Chris A. Janiszewski. 1989. “Assessing the Role ofContingency Awareness in Attitudinal Conditioning With Implica-tions for Advertising Research.”Journal of Marketing Research26(February): 30-43.

Austin, James T. and Jeffrey B. Vancouver. 1996. “Goal Constructs inPsychology: Structure, Process, and Content.”Psychological Bulle-tin 120:338-375.


Bagozzi, Richard P. 1991a. “Further Thoughts on the Validity of Mea-sures of Elation, Gladness, and Joy.”Journal of Personality and So-cial Psychology61:98-104.

. 1991b. “The Role of Psychophysiology in Consumer Research.”In Handbook of Consumer Behavior. Eds. Thomas S. Robertson andHarold H. Kassarjian. Englewood Cliffs, NJ: Prentice Hall, 124-161.

. 1992. “The Self-Regulation of Attitudes, Intentions, and Behav-ior.” Social Psychology Quarterly55:178-204.

. 1993. “An Examination of the Psychometric Properties ofMeasures of Negative Affect in the PANAS-X Scales.”Journal ofPersonality and Social Psychology65:836-851.

. 1994. “The Effects of Arousal on the Organization of Positiveand Negative Affect and Cognitions: Application to Attitude The-ory.” Structural Equation Modeling1:222-252.

. 1996. “The Role of Arousal in the Creation and Control of theHalo Effect in Attitude Models.”Psychology & Marketing13:235-264.

. Forthcoming. “Happiness.” InEncyclopedia of Human Emo-tions. Eds. David Levinson, James Ponzetti, and Peter Jorgensen.New York: Macmillan.

and Robert E. Burnkrant. 1979. “Attitude Organization and theAttitude-Behavior Relationship.”Journal of Personality and SocialPsychology37:913-929.

and Utpal Dholakia. 1998. “Goal-Setting and Goal-Striving inConsumer Behavior.” Unpublished working paper, University ofMichigan.

and Elizabeth A. Edwards. 1998. “Goal Setting and Goal Pursuitin the Regulation of Body Weight.”Psychology and Health13:593-621.

and David J. Moore. 1994. “Public Service Advertisements:Emotion and Empathy Guide Prosocial Behavior.”Journal of Mar-keting58:56-70.

, Hans Baumgartner, and Rik Pieters. 1998. “Goal-Directed Emo-tions.” Cognition and Emotion12:1-26.

, Nancy Wong, and Youjae Yi. 1998. “The Representation of Af-fect in Independent- and Interdependent-Based Cultures.” Unpub-lished working paper, University of Michigan.

Barrett, Lisa Feldman and James A. Russell. 1998. “Independence andBipolarity in the Structure of Current Affect.”Journal of Personalityand Social Psychology74:967-984.

Bartlett, James C. and John W. Santrock. 1979. “Affect-Dependent Epi-sodic Memory in Young Children.”Child Development50:513-518.

, Georgia Burleson, and John W. Santrock. 1982. “EmotionalMood and Memory in Young Children.”Journal of ExperimentalChild Psychology34:59-76.

Batra, Rajeev and Olli T. Ahtola. 1990. “Measuring the Hedonic andUtilitarian Sources of Consumer Attitudes.”Marketing Letters2:159-170.

and Morris B. Holbrook. 1990. “Developing a Typology of Af-fective Responses to Advertising.”Psychology & Marketing7(Spring): 11-25.

and Michael L. Ray. 1986. “Affective Responses Mediating Ac-ceptance of Advertising.”Journal of Consumer Research13 (Sep-tember): 234-249.

and Douglas M. Stayman. 1990. “The Role of Mood in Advertis-ing Effectiveness.”Journal of Consumer Research17:203-214.

and Debra Stephens. 1994. “Attitudinal Effects of Ad-EvokedMoods and Emotions: The Moderating Role of Motivation.”Psychol-ogy & Marketing11:199-215.

Bless, Herbert, Gerd Bohner, Norbert Schwarz, and Fritz Strack. 1990.“Mood and Persuasion: A Cognitive Response Analysis.”Personal-ity and Social Psychology Bulletin16 (2): 331-345.

, Gerald L. Clore, Norbert Schwarz, Verena Golisano, ChristinaRabe, and Marcus Wolk. 1996. “Mood and the Use of Scripts: Does aHappy Mood Really Lead to Mindlessness?”Journal of Personalityand Social Psychology71 (4): 665-679.

Bohner, Gerd, Kimberly Crow, Hans-Peter Erb, and Norbert Schwarz.1992. “Affect and Persuasion: Mood Effects on the Processing ofMessage Content and Context Cues and on Subsequent Behaviour.”European Journal of Social Psychology22:511-530.

, Shelly Chaiken, and Piroska Hunyadi. 1994. “The Role of Moodand Message Ambiguity in the Interplay of Heuristic and SystematicProcessing.”European Journal of Social Psychology24:207-221

Boninger, David S., Faith Gleicher, John Hetts, David Armor, and E.Moore. 1994. “The Influence of Counterfactual Thinking on Inten-tions and Behavior.” Unpublished raw data, reported in Gleicher et al.(1995).

Bower, Gordan H. 1981. “Mood and Memory.”American Psychologist36 (2): 129-148.

and Paul R. Cohen. 1982. “Emotional Influences in Memory andThinking: Data and Theory.” InAffect and Cognition. Eds. MargaretClark and Susan T. Fiske. Hillsdale, NJ: Lawrence Erlbaum, 291-331.

and John D. Mayer. 1985. “Failure to Replicate Mood-Dependent Retrieval.”Bulletin of the Psychonomic Society23:39-42.

, Kenneth P. Monteiro, and Stephen G. Gilligan. 1978. “Emo-tional Mood as a Context for Learning and Recall.”Journal of VerbalLearning and Verbal Behavior17:573-585.

, Stephen G. Gilligan, and Kenneth P. Monteiro. 1981. “Selectiv-ity of Learning Caused by Affective States.”Journal of ExperimentalPsychology: General110 (4): 451-473.

Bradburn, Norman M. 1969.The Structure of Psychological Well-Being.Chicago: Aldine.

Breckler, Steven J. and Elizabeth C. Wiggins. 1989. “Affect VersusEvaluation in the Structure of Attitudes.”Journal of ExperimentalSocial Psychology25:253-271.

Brown, Steven P. and Douglas M. Stayman. 1992. “Antecedents andConsequences of Attitude Toward the Ad: A Meta-Analysis.”Jour-nal of Consumer Research19 (June): 34-51.

, William L. Cron, and John W. Slocum Jr. 1997. “Effects of Goal-Directed Emotions on Salesperson Volitions, Behavior, and Perfor-mance: A longitudinal Study.”Journal of Marketing61:39-50.

Burke, Marian C. and Julie A. Edell. 1989. “The Impact of Feelings onAd-Based Affect and Cognition.”Journal of Marketing Research26:69-83.

Cacioppo, John T., Gary G. Berntson, and David J. Klein. 1992. “What Isan Emotion? The Role of Somatovisceral Afference, With SpecialEmphasis on Somatovisceral ‘Illusions.’ ”Review of Personality andSocial Psychology14:63-98.

Cantor, Joanne R., Jennings Bryant, and Dolf Zillman. 1974. “Enhance-ment of Humor Appreciation by Transferred Excitation.”Journal ofPersonality and Social Psychology30:812-821.

Clark, Margaret D. and John D. Teasdale. 1982. “Diurnal Variation inClinical Depression and Accessibility of Memories of Positive andNegative Experiences.”Journal of Abnormal Psychology91:87-95.

Clore, Gerald L. and Donn Byrne. 1974. “A Reinforcement AffectModel ofAttraction.” In Foundations of Interpersonal Attraction.Ed. T. L. Huston. New York: Academic Press, 143-170.

, Norbert Schwarz, and Michael Conway. 1994. “AffectiveCauses and Consequences of Social Information Processing.” InHandbook of Social Cognition, 2nd ed., Vol. 1. Eds. Robert Wyer,Thomas Srull, and Alice Isen. Hillsdale, NJ: Lawrence Erlbaum,323-417.

Cohen, Joel B. and Charles S. Areni. 1991. “Affect and Consumer Behav-ior.” In Handbook of Consumer Behavior. Eds. Thomas S. Robertsonand Harold H. Kassarjian. Englewood Cliffs, NJ: Prentice Hall, 188-240.

Crites Jr., Stephen L., Leandre R. Fabrigar, and Richard E. Petty. 1994.“Measuring the Affective and Cognitive Properties of Attitudes:Conceptual and Methodological Issues.”Personality and Social Psy-chology Bulletin20 (December): 619-634.

Damasio, Antonio R. 1994.Descartes’Error: Emotion, Reason, and theHuman Brain. New York: Avon.

Diener, Ed and Robert A. Emmons 1985. “The Independence of Positiveand Negative Affect.”Journal of Personality and Social Psychology47:1105-1117.

and Randy J. Larsen. 1993. “The Experience of Emotional Well-Being.” InHandbook of Emotions. Eds. Michael Lewis and JeannetteM. Haviland. New York: Guilford, 405-415.

Donovan, Robert J. and John R. Rossiter. 1982. “Store Atmospherics:An Environmental Psychology Approach.”Journal of Retailing58:34-57.


Dubé, Laurette, Marie-Claude Bélanger, and Elyse Trudeau. 1996. “TheRole of Emotions in Health Care Satisfaction.”Journal of HealthCare Marketing16 (Summer): 45-51.

Dutton, Donald G. and Arthur P. Aron. 1974. “Some Evidence forHeightened Sexual Attraction Under Conditions of High Anxiety.”Journal of Personality and Social Psychology30:510-517.

Eagly, Alice H. and Shelly Chaiken. 1993.The Psychology of Attitudes.New York: Harcourt Brace Jovanovich.

, Antonio Mladinic, and Stacey Otto. 1994. Cognitive and Affec-tive Bases of Attitudes Toward Social Groups and Social Policies.”Journal of Experimental Social Psychology30:113-137.

Edell, Julie A. and Marian Chapman Burke. 1987. “The Power of Feel-ings in Understanding Advertising Effects.”Journal of ConsumerResearch14 (December): 421-433.

Eich, Eric J. and Isabel M. Birnbaum. 1982. “Repetition, Cueing andState-Dependent Memory.”Memory and Cognition10:103-114.

Fiedler, Klaus. 1990. “Mood-dependent Selectivity in Social Cognition.”In European Review of Social Psychology, Vol. 1. Eds. Wolfgang Stroebeand Miles Hewstone. New York: John Wiley, 1-32.

Fishbein, Martin and Icek Ajzen. 1975.Belief, Attitude, Intention, andBehavior: An Introduction to Theory and Research. Reading, MA:Addison-Wesley.

Folkes, Valerie, Susan Koletsky, and John L. Graham. 1987. “A FieldStudy of Causal Inferences and Consumer Reaction: The View Fromthe Airport.” Journal of Consumer Research13 (March): 534-539.

Forgas, Joseph P. 1994. “The Role of Emotion in Social Judgments: AnIntroductory Review and an Affect Infusion Model (AIM).”Euro-pean Journal of Social Psychology24:1-24.

. 1995. “Mood and Judgment: The Affect Infusion Model(AIM).” Psychological Bulletin117 (1): 39-66.

and Gordon H. Bower. 1987. “Mood Effects on Person-Perception Judgments.”Journal of Personality and Social Psychol-ogy53 (1): 53-60.

Friestad, Marian and Esther Thorson. 1986. “Emotion-Eliciting Adver-tising: Effect on Long-term Memory and Judgment.”Advances inConsumer Research13:111-115.

Frijda, Nico H. 1986.The Emotions. Cambridge, UK: Cambridge Uni-versity Press.

. 1993. “Moods, Emotion Episodes, and Emotions.” InHandbookof Emotions. Eds. M. Lewis and J. M. Haviland. New York: Guilford,381-403.

. 1994. “Emotions Are Functional, Most of the Time.” InThe Na-ture of Emotion. Eds. P. Ekman and R. J. Davidson. New York: Ox-ford University Press, 112-122.

, Peter Kuipers, and Elisabeth ter Schure. 1989. “RelationsAmong Emotion, Appraisal, and Emotional Action Readiness.”Journal of Personality and Social Psychology57:212-228.

Gardner, Meryl P. 1985. “Mood States and Consumer Behavior: A Criti-cal Review.”Journal of Consumer Research12 (December): 281-300.

and G. J. Simokos. 1986. “Toward a Methodology for AssessingEffects of Instore Atmospherics.”Advances in Consumer Research13:27-31.

Gleicher, Faith, David S. Boninger, Alan Strathman, David Armor, JohnHetts, and Mina Ahn. 1995. “With an Eye Toward the Future: The Im-pact of Counterfactual Thinking on Affect, Attitudes, and Behavior.”In What Might Have Been: The Social Psychology of CounterfactualThinking. Eds. Neal J. Roese and James M. Olson. Mahwah, NJ:Lawrence Erlbaum, 283-304.

Goldberg, Marvin E. and Gerald J. Gorn. 1987. “Happy and Sad TV Pro-grams: How They Affect Reactions to Commercials.”Journal ofConsumer Research14:387-403.

Gopinath, Mahesh and Richard P. Bagozzi. 1999. “Cognitive AppraisalsLeading to Consumer Emotions and Action Tendencies.” Unpub-lished working paper, University of Michigan.

Green, Donald P., Susan L. Goldman, and Peter Salovey.1993. “Mea-surement Error Masks Bipolarity in Affect Ratings.”Journal of Per-sonality and Social Psychology64:1029-1041.

Hayes-Roth, B. 1977. “Evolution of Cognitive Structures and Pro-cesses.”Psychological Review84:260-278.

Hill, Ronald P. and Meryl P. Gardner. 1987. “The Buying Process: Effectsof and on Consumer Mood States.”Advances in Consumer Research14:408-410.

Hochschild, Arlie R. 1983.The Managed Heart. Berkeley: University ofCalifornia Press.

Holbrook, Morris B. and Rajeev Batra. 1987. “Assessing the Role ofEmotions as Mediators of Consumer Responses to Advertising.”Journal of Consumer Research14:404-420.

Isen, Alice M. 1984. “Toward Understanding the Role of Affect in Cog-nition.” In Handbook of Social Cognition, Vol. 3. Eds. Robert Wyer,Thomas Srull, and Alice Isen. Hillsdale, NJ: Lawrence Erlbaum, 179-236.

. 1987. “Positive Affect, Cognitive Processes, and Social Behav-ior.” In Advances in Experimental Social Psychology, Vol. 20. Ed.Leonard Berkowitz. New York: Academic Press, 203-253.

. 1989. “Some Ways in Which Affect Influences Cognitive Pro-cesses: Implications for Advertising and Consumer Behavior.” InCognitive and Affective Responses to Advertising. Eds. Patricia Caf-ferata and Alice M. Tybout. Lexington, MA: Lexington Books, 91-117.

——— and Kimberly A. Daubman. 1984. “The Influence of Affect onCategorization.”Journal of Personality and Social Psychology47(6): 1206-1217.

, Thomas E. Shalker, Margaret Clark, and Lynn Karp. 1978. “Af-fect, Accessibility of Material in Memory, and Behavior: A CognitiveLoop?”Journal of Personality and Social Psychology36 (1): 1-12.

, Mitzi M. Johnson, Elizabeth Mertz, and Gregory F. Robinson.1985. “The Influence of Positive Affect on the Unusualness of WordAssociations.”Journal of Personality and Social Psychology48 (6):1-14.

, Kimberly A. Daubman, and Gary P. Nowicki. 1987. “PositiveAffect Facilitates Creative Problem Solving.”Journal of Personalityand Social Psychology52 (6): 1122-1131.

, Paula M. Niedenthal, and Nancy Cantor. 1992. “An Influence ofPositive Affect on Social Categorization.”Motivation and Emotion16 (1): 65-78.

Izard, Carroll E. 1972.Patterns of Emotions: A New Analysis of Anxietyand Depression. New York: Academic Press.

. 1991.The Psychology of Emotions. New York: Plenum.. 1992. “Basic Emotions, Relations Among Emotions, and

Emotion-Cognition Relations.”Psychological Review99:561-565.James, William. [1890] 1950.The Principles of Psychology. Vol 2. New

York: Holt.Laird, James D., John J. Wagener, Mark Halal, and Martha Szegda. 1982.

“Remembering What You Feel: The Effects of Emotion on Memory.”Journal of Personality and Social Psychology42 (4): 646-657.

Larsen, Randy J. and Ed Diener. 1992. “Promises and Problems With theCircumplex Model of Emotion.” InReview of Personality and SocialPsychology: Vol. 14. Emotional and Social Behavior. Ed. M. S.Clark. Newbury Park, CA: Sage, 25-59.

Lazarus, Richard S. 1982. “Thoughts on the Relations Between Emotionand Cognition.”American Psychologist35:1019-1024.

. 1991.Emotion and Adaptation. New York: Oxford UniversityPress.

LeDoux, Joseph. 1996.The Emotional Brain: The Mysterious Underpin-nings of Emotional Life. New York: Simon & Schuster.

Lewis, Michael 1993. “Self-Conscious Emotions: Embarrassment,Pride, Shame, and Guilt.” InHandbook of Emotions. Eds. M. Lewisand J. M. Haviland. New York: Guilford, 563-573.

Locke, Karen. 1996. “A Funny Thing Happened! The Management ofConsumer Emotions in Service Encounters.”Organizational Science7 (January-February): 40-59.

MacKenzie, Scott B., Richard J. Lutz, and George Belch. 1986. “TheRole of Attitude Toward the Ad as a Mediator of Advertising Effec-tiveness: A Test of Competing Explanations.”Journal of MarketingResearch23 (May): 130-143.

Mackie, Diane M. and Leila T. Worth. 1989. “Processing Deficits and theMediation of Positive Affect in Persuasion.”Journal of Personalityand Social Psychology57 (1): 27-40.

Mano, Haim and Richard L. Oliver. 1993. “Assessing the Dimensionalityand Structure of the Consumption Experience: Evaluation, Feeling,


and Satisfaction.”Journal of Consumer Research20 (December):451-466.

Manstead, Antony S. R. and Hugh L. Wagner. 1981. “Arousal, Cognition,and Emotion: An Appraisal of Two-Factor Theory.”Current Psycho-logical Reviews1: 35-54.

Marcel, Anthony J. 1983. “Conscious and Unconscious Perception: AnApproach to the Relations Between Phenomenal Experience and Per-ceptual Processes.”Cognitive Psychology15:238-300.

Markus, Hazel R. and Shinobu Kitayama. 1994. “The Cultural Con-struction of Self and Emotion: Implications for Social Behavior.” InEmotion and Culture: Empirical Studies of Mutual Influence. Eds.S. Kitayama and H. R. Markus. Washington, DC: American Psycho-logical Association, 89-130.

Martin, Leonard L., David W. Ward, John W. Achee, and Robert S. Wyer Jr.1993. “Mood as Input: People Have to Interpret the Motivational Im-plications of Their Moods.”Journal of Personality and Social Psy-chology64 (3): 317-326.

Milliman, Ronald E. 1982. “Using Background Music to Affect the Be-havior of Supermarket Shoppers.”Journal of Marketing46 (Sum-mer): 86-91.

. 1986. “The Influence of Background Music on the Behavior ofRestaurant Patrons.”Journal of Consumer Research13:286-289.

Miniard, Paul W., Sunil Bhatla, and Randall L. Rose. 1990. “On the For-mation and Relationship of Ad and Brand Attitudes: An Experimen-tal and Causal Analysis.”Journal of Marketing Research27(August): 290-303.

Mitchell, Andrew A. and Jerry C. Olson. 1981. “Are Product AttributeBeliefs the Only Mediator of Advertising Effects on Brand Atti-tude?”Journal of Marketing Research18 (August): 318-332.

Murray, Noel, Harish Sujan, Edward R. Hirt, and Mita Sujan. 1990. “TheInfluence of Mood on Categorization: A Cognitive Flexibilty Inter-pretation.”Journal of Personality and Social Psychology59 (3): 411-425.

Nasby, William and Regina Yando. 1982. “Selective Encoding and Re-trieval of Affectively Valent Information.”Journal of Personality andSocial Psychology43 (6): 1244-1253.

Nyer, Prashanth U. 1997a. “Modeling the Cognitive Antecedents of Post-Consumption Emotions.”Journal of Consumer Satisfaction, Dissat-isfaction and Complaining Behavior10:80-90.

. 1997b. “A Study of the Relationships Between Cognitive Ap-praisals and Consumption Emotions.”Journal of the Academy ofMarketing Science25 (4): 296-304.

Oatley, Keith 1992.Best Laid Schemes: The Psychology of Emotions.Cambridge, UK: Cambridge University Press.

and Philip N. Johnson-Laird. 1987. “Towards a Cognitive Theoryof Emotions.”Cognition and Emotion1:29-50.

Obermiller, Carl. 1985. “Varieties of Mere Exposure: The Effects ofProcessing Style and Repetition on Affective Response.”Journal ofConsumer Research12:17-30.

Oliver, Richard L. 1980. “A Cognitive Model of the Antecedents andConsequences of Satisfaction Decisions.”Journal of Marketing Re-search17 (November): 460-469.

. 1993. “Cognitive, Affective, and Attribute Bases of the Satisfac-tion Response.”Journal of Consumer Research20 (December): 418-430.

. 1994. “Conceptual Issues in the Structural Analysis of Con-sumption Emotion, Satisfaction, and Quality.” InAdvances in ConsumerResearch, Vol. 21. Eds. Chris T. Allen and Deborah Roedder John.Provo, UT: Association for Consumer Research, 16-22.

Olney, Thomas J., Morris B. Holbrook, and Rajeev Batra. 1991. “Con-sumer Responses to Advertising: The Effects of Ad Content, Emo-tions, and Attitude Toward the Ad on Viewing Time.”Journal ofConsumer Research17 (March): 440-453.

Ortony, Andrew, Gerald L. Clore, and Allan Collins. 1988.The CognitiveStructure of Emotions. Cambridge, UK: Cambridge University Press.

Parker, Dianne, Antony S. R. Manstead, and Stephen G. Stradling. 1995.“Extending the Theory of Planned Behaviour: The Role of PersonalNorm.” British Journal of Social Psychology34:127-137.

Parkinson, Brian. 1995.Ideas and Realities of Emotion. London:Routledge.

Parrott, Gerrod W. 1988. “The Role of Cognition in Emotional Experi-ence.” InRecent Trends in Theoretical Psychology. Eds. W. J. Baker,

L. P. Mos, H. V. Rappard, and H. J. Stam. New York: Springer-Verlag,327-337.

Perugini, Marco and Richard P. Bagozzi. 1998. “The Role of Desires andAnticipatory Emotions in Goal-Directed Behaviors: Extending theTheory of Planned Behavior.” Unpublished working paper, Univer-sity of Michigan Business School.

Petty, Richard E. and John T. Cacioppo. 1986.Communication and Per-suasion. New York: Springer-Verlag.

, David W. Schumann, Steven A. Richman, and Alan J. Strathman.1993. “Positive Mood and Persuasion: Different Roles for Affect Un-der High and Low Elaboration Conditions.”Journal of Personalityand Social Psychology64:5-20.

Pieters, Rik, Hans Baumgartner, and D. Allen. 1995. “A Means-EndChain Approach to Consumer Goal Structures.”International Jour-nal of Research in Marketing12 (3): 227-244.

Plutchik, Robert 1980. Emotion:A Psychoevolutionary Synthesis. NewYork: Harper & Row.

Puto, Christopher P. and William D. Wells. 1984. “Informational andTransformational Advertising: The Differential Effects of Time.”Ad-vances in Consumer Research11:638-643.

Reisenzein, Rainer 1983. “The Schachter Theory of Emotion: Two Dec-ades Later.”Psychological Bulletin94:239-264.

Richard, René, Joop van der Pligt, and Nanne de Vries. 1995. “Antici-pated Affective Reactions and Prevention of AIDS.”British Journalof Social Psychology34:9-21.

Richins, Marsha L. 1997. “Measuring Emotions in the Consumption Ex-perience.”Journal of Consumer Research24:127-146.

Roese, Neal J. and James M. Olson, eds. 1995.What Might Have Been:The Social Psychology of Counterfactual Thinking. Mahwah, NJ:Lawrence Erlbaum.

Roseman, Ira J. 1991. “Appraisal Determinants of Discrete Emotions.”Cognition and Emotion5:161-200.

Rozin, Paul, Jonathan Haidt, and Clark R. McCauley. 1993. “Disgust.” InHandbook of Emotions. Eds. M. Lewis and J. M. Haviland. NewYork: Guilford, 575-594.

Russell, James A. 1980. “A Circumplex Model of Affect.”Journal ofPersonality and Social Psychology39 (December): 1161-1178.

. 1997. “How Shall an Emotion Be Called.” InCircumplex Mod-els of Personality and Emotions. Eds. R. Plutchick and H. R. Conte.Washington, DC: American Psychological Association, 205-220.

and Albert Mehrabian. 1977. “Evidence for a Three-Factor The-ory of Emotions.”Journal of Research in Personality11:273-294.

, Anna Weiss, and Gerald A. Mendelsohn. 1989. “Affect Grid: ASingle-Item Scale of Pleasure and Arousal.”Journal of Personalityand Social Psychology57:493-502.

Ruth, Julie A., Cele C. Otnes, and Frédéric F. Brunel. 1998. “Bonds andBurdens: Gift Receipt and Relationship Reformulation.” Unpub-lished working paper, University of Washington.

Salovey,Peter and Deborah Birnbaum. 1989. “Influence of Mood onHealth-Related Cognitions.”Journal of Personality and Social Psy-chology57:539-551.

Sanbonmatsu, David M. and Frank R. Kardes. 1988. “The Effect ofPhysiological Arousal on Information Processing and Persuasion.”Journal of Consumer Research15:379-395.

Schachter, Stanley and Jerome E. Singer. 1962. “Cognitive, Social, andPhysiological Determinants of Emotional State.”Psychological Re-view69:379-399.

Schaller, Mark and Robert B. Cialdini. 1990. “Happiness, Sadness, andHelping: A Motivational Integration.” InHandbook of Motiva-tion and Cognition: Foundations of Social Behavior, Vol. 2. Eds.R. M. Sorrentino and E. T. Higgins. New York: Guilford, 265-296.

Schwarz, Norbert. 1990. “Feelings as Information: Informational andMotivational Functions of Affective States.” InHandbook of Motiva-tion and Cognition: Foundations of Social Behavior, Vol. 2. Eds.R. M. Sorrentino and E. T. Higgins. New York: Guilford, 527-561.

and Herbert Bless. 1991. “Happy and Mindless, But Sad andSmart? The Impact of Affective States on Analytic Reasoning.” InEmotion and Social Judgments. Ed. Joseph P. Forgas. London: Perga-mon, 55-71.

and Gerald L. Clore. 1983. “Mood, Misattribution, and Judg-ments of Well-Being: Informative and Directive Functions of Affec-


tive States.”Journal of Personality and Social Psychology45 (3):513-523.

, Fritz Strack, Detlev Kommer, and Dirk Wagner. 1987. “Soccer,Rooms and the Quality of Your Life: Mood Effects on Judgments ofSatisfaction With Life in General and with Specific Life Domains.”European Journal of Social Psychology17:69-79.

Shaver, Phillip, Judith Schwartz, Donald Kirson, and Cary O’Connor.1987. “Emotion Knowledge: Further Exploration of a Prototype Ap-proach.” Journal of Personality and Social Psychology52:1061-1086.

Shimp, Terence A. 1981. “Attitude Toward the Ad as a Mediator of Con-sumer Brand Choice.”Journal of Advertising10:9-15.

. 1991. “Neo-Pavlovian Conditioning and Its Implications forConsumer Theory and Research.” InHandbook of Consumer Behav-ior. Eds. Thomas S. Robertson and Harold H. Kassarjian. EnglewoodCliffs, NJ: Prentice Hall, 162-187.

Smith, Craig A. and Phoebe C. Ellsworth. 1985. “Patterns of CognitiveAppraisals in Emotion.”Journal of Personality and Social Psychol-ogy48:813-838.

Srull, Thomas K. 1983. “Affect and Memory: The Impact of AffectiveReactions in Advertising on the Representation of Product Informa-tion in Memory.” InAdvances in Consumer Research, Vol. 10. Eds.Richard P. Bagozzi and Alice Tybout. Ann Arbor, MI: Associationfor Consumer Research, 520-525.

. 1984. “The Effects of Subjective Affective States on Memoryand Judgment.” InAdvances in Consumer Research, Vol. 11. Ed.Thomas C. Kinnear. Provo, UT: Association for Consumer Research,530-533.

Stayman, Douglas M. and David A. Aaker. 1988. “Are All the Effects ofAd-Induced Feelings Mediated by Aad?”Journal of Consumer Re-search15:368-373.

and Rajeev Batra. 1991. “Encoding and Retrieval of Ad Affect inMemory.” Journal of Marketing Research28 (May): 232-239.

Stein, Nancy L., Maria D. Liwag, and Elizabeth Wade. 1996. “A Goal-Based Approach to Memory for Emotional Events: Implementationsfor Theories of Understanding and Socialization.” InEmotion: Inter-disciplinary Perspectives. Eds. R. D. Kavanaugh, B. Zimmerberg,and S. Fein. Mahwah, NJ: Lawrence Erlbaum, 91-118.

Taylor, Shirley. 1994. “Waiting for Service: The Relationship BetweenDelays and Evaluations of Services.”Journal of Marketing58(April): 56-69.

Teasdale, John D. and Sarah J. Fogarty. 1979. “Differential Effects of In-duced Mood on Retrieval of Pleasant and Unpleasant Events FromEpisodic Memory.”Journal of Abnormal Psychology88 (3): 248-257.

and Louise M. Russell. 1983. “Differential Aspects of InducedMood on the Recall of Positive, Negative and Neutral Words.”BritishJournal of Clinical Psychology22:163-171.

Tulving, Endel and Zena Pearlstone. 1966. “Availability Versus Accessi-bility of Information in Memory for Words.”Journal of VerbalLearning and Verbal Behavior5:381-391.

Verbeke, Willem and Richard P. Bagozzi. 1998. “Self-Conscious Emo-tions in the Context of Personal Selling: Performance Is a Function ofthe Ability to Cope With Shame and Embarrassment.” Unpublishedworking paper, University of Michigan.

and. 1999. “Coping With Anxiety in Personal Selling: AStudy of Financial Service Salespeople.” Unpublished working pa-per, University of Michigan.

Watson, David and Auke Tellegen. 1985. “Toward a Consensual Struc-ture of Mood.”Psychological Bulletin98 (September): 219-235.

Wegener, Duane T. and Richard E. Petty. 1994. “Mood-ManagementAcross Affective States: The Hedonic Contingency Hypothesis.”Journal of Personality and Social Psychology66 (6): 1034-1048.

and. 1996. “Effects of Mood on Persuasion Processes:Enhancing, Reducing, and Biasing Scrutiny of Attitude-Relevant In-formation.” InStriving and Feeling: Interactions Between Goals andAffect. Eds. L. L. Martin and A. Tesser. Mahwah, NJ: Lawrence Erl-baum, 329-362.

, , and David J. Klein. 1994. “Effects of Mood on HighElaboration Attitude Change: The Mediating Role of LikelihoodJudgments.”European Journal of Social Psychology24:25-43.

, , and Stephen M. Smith. 1995. “Positive Mood Can In-crease or Decrease Message Scrutiny: The Hedonic ContingencyView of Mood and Message Processing.”Journal of Personality andSocial Psychology69 (1): 5-15.

Westbrook, Robert A. 1987. “Product/Consumption-Based AffectiveResponses and Postpurchase Processes.”Journal of Marketing Re-search24 (August): 258-270.

and Richard L. Oliver. 1991. “The Dimensionality of Consump-tion Emotion Patterns and Consumer Satisfaction.”Journal of Con-sumer Research18 (June): 84-91.

Wiles, Judith A. and T. Bettina Cornwell. 1990. “A Review of MethodsUtilized in Measuring Affect, Feelings, and Emotion in Advertising.”Current Issues and Research in Advertising13:241-275.

Worth, Leila T. and Diane M. Mackie. 1987. “Cognitive Mediation ofPositive Affect in Persuasion.”Social Cognition5:76-94.

Yi, Youjae. 1990. “A Critical Review of Consumer Satisfaction.” InRe-view of Marketing 1990. Ed. V. A. Zeithaml. Chicago: AmericanMarketing Association, 68-123.

Zajonc, Robert B. 1968. “Attitudinal Effects of Mere Exposure.”Journalof Personality and Social Psychology Monograph Supplement9:1-27.

. 1980. “Feeling and Thinking: Preferences Need No Inferences.”American Psychologist35:151-175.

. 1984. “On the Primacy of Affect.”American Psychologist39:117-123.

. 1998. “Emotions.” InThe Handbook of Social Psychology. Eds.D. T. Gilbert, S. T. Fiske, and G. Lindzey. Boston: McGraw-Hill, 591-632.

and Hazel Markus. 1982. “Affective and Cognitive Factors inPreferences.”Journal of Consumer Research9 (September): 123-131.

Zevon, Michael A. and Auke Tellegen. 1982. “The Structure of MoodChange: An Idiographic/Nomothetic Analysis.”Journal of Personal-ity and Social Psychology43:111-122.

Zillman, Dolf. 1971. “Excitation Transfer in Communication-MediatedAggressive Behavior.”Journal of Experimental Social Psychology7:419-434.

. 1983. “Transfer of Excitation in Emotional Behavior.” InHand-book of Social Psychology. Eds. John T. Cacioppo and Richard E.Petty. New York: Guilford, 215-230.

ABOUT THE AUTHORS

Richard P. Bagozziis the Dwight F. Benton Professor of Behav-ioral Science in Management at the University of Michigan Busi-ness School. He is a graduate of Northwestern University and hasbeen a faculty member at the University of California–Berkeley,the Massachusetts Institute of Technology, and Stanford Univer-sity. He is currently doing research in emotions, goal-directed be-havior, and social identity theory.

Mahesh Gopinath is an assistant professor of marketing in theA. B. Freeman School of Business at Tulane University. He is agraduate of the University of Michigan Business School and isdoing research in emotions and customer satisfaction.

Prashanth U. Nyer is an assistant professor of marketing in theSchool of Business and Economics at Chapman University. He isa graduate of the University of Michigan Business School and isdoing research in emotions and customer satisfaction.


Documents

The Role of Emotions in Marketing