Valenced Emotions in Satisfaction: a Look At Affect in Shopping

Mai Neo, University of Pittsburgh
Audrey J. Murrell, University of Pittsburgh
ABSTRACT - Emotion and satisfaction were studied in the context of shopping for oneself and for another person. Four clusters were identified for each situation: (1) positive/happy, (2) unemotional, (3) negative/sad and (4) positive/efficacy. Three core emotions were also identified, (1) efficacy, (2) negative and (3) positive while shopping for oneself, and (1) efficacy and (2) negative while shopping for others. The results showed that people have positive/happy emotions while shopping for themselves and positive/efficacious emotions while shopping for others. This positive/happy cluster exhibited higher satisfaction while shopping for oneself, whereas the positive/efficacy cluster exhibited higher satisfaction while shopping for other people.
[ to cite ]:
Mai Neo and Audrey J. Murrell (1993) ,"Valenced Emotions in Satisfaction: a Look At Affect in Shopping", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 667-672.

Advances in Consumer Research Volume 20, 1993      Pages 667-672

VALENCED EMOTIONS IN SATISFACTION: A LOOK AT AFFECT IN SHOPPING

Mai Neo, University of Pittsburgh

Audrey J. Murrell, University of Pittsburgh

ABSTRACT -

Emotion and satisfaction were studied in the context of shopping for oneself and for another person. Four clusters were identified for each situation: (1) positive/happy, (2) unemotional, (3) negative/sad and (4) positive/efficacy. Three core emotions were also identified, (1) efficacy, (2) negative and (3) positive while shopping for oneself, and (1) efficacy and (2) negative while shopping for others. The results showed that people have positive/happy emotions while shopping for themselves and positive/efficacious emotions while shopping for others. This positive/happy cluster exhibited higher satisfaction while shopping for oneself, whereas the positive/efficacy cluster exhibited higher satisfaction while shopping for other people.

BACKGROUND

Emotion can be thought of as a specific instance of affect states. Emotions are an intense and stimulus-specific affect state (Clark and Isen 1982; Gardner 1985) that demand more attention to a stimulus and can disrupt ongoing goal-directed activity (Clark and Isen 1982). To the extent that emotions appear in marketing research, they tend to be conceptualized in a very global context, where the individual components are aggregated to form a single positive-negative emotion continuum (Schwartz and Shaver 1987; Shaver, Schwartz, Kirson and O'Connor 1987; Westbrook 1987).

In marketing, emotion is extensively studied in the advertising and media contexts in terms of positive or negative feelings toward the ad (i.e., prepurchase situations). It has been noted by researchers that exploring the antecedents of attitudes, such as emotions, is important because they are responsible for the consumer's liking or disliking of the ad (Murry, Lastovicka and Singh 1992), act as intervening variables that mediate the relationship between ad content and viewing time (Olney, Holbrook and Batra 1991) and have strong effects on attitude toward the ad (Holbrook and Batra 1987; MacInnis and Park 1991). Consumers frequently generate more than one emotion after exposures to stimuli (Polivy 1981). For example, Aaker, Stayman and Vezina (1988) provided an inventory of 31 feeling clusters, 16 positive and 15 negative that were elicited by advertising (e.g., "delighted", ""warm/tender", and "sad"). However, these prepurchase situations are not the only context where emotions play an important role. Other researchers have explored the impact of affect in postpurchase contexts. Research in this area of marketing has looked at the impact on consumer satisfaction and product consumption (Westbrook 1980; Westbrook 1987; Westbrook and Oliver 1991). This research asserts that satisfaction is not solely a cognitive phenomenon. Rather, it is also comprised of affect or feeling, in that consumers feel subjectively good (satisfaction), or subjectively bad (dissatisfaction) concerning shopping decisions. Westbrook (1980) demonstrated that affect is linked to high levels of consumer satisfaction. While this research assumes that multiple sources of affect produce an effect on product satisfaction, limited evidence exists that specifies a taxonomy for these multiple affect sources.

Recently, Westbrook and Oliver (1991) investigated the interrelationship between consumption emotion and satisfaction by way of taxonomic and dimensional analyses to identify patterns of emotional response to product experiences. Consumption emotion refers to emotional responses elicited during product usage or consumption experiences. These emotions can be can be assessed as distinctive categories (e.g., joy, anger, and fear) or they can be analyzed in terms of structural dimensions underlying these emotional categories, such as pleasantness/unpleasantness, relaxation/action, or calmness/excitement (Russell 1979).

To develop this taxonomy, Westbrook and Oliver (1991) conducted a field study on a sample of owners of newly purchased cars. A convenience sample of 125 respondents were surveyed on their "feelings and attitudes" toward their most recent car purchase. To measure consumption emotion, Izard's (1977) DES-II scale was used. Positive emotions examined were interest and joy. Negative emotions surveyed were anger, contempt, disgust, shame, guilt, sadness, and fear. Surprise, as a neutral emotion, was also examined along with an overall consumer satisfaction scale.

Using a k-means cluster analysis, a five-cluster solution was revealed. The happy/content cluster contained people who reported frequent interest and joy but infrequent surprise and negative emotions. Those in the pleasant/ surprise cluster were people who were high on joy and surprise but low on all negative emotions. The unemotional cluster contained subjects whose scores fell below all measures of consumption emotion, especially joy and surprise. Subjects in the unpleasant/surprise cluster were high on surprise and most negative emotions (especially sadness) yet low on joy. The angry/upset cluster contained subjects who reported frequent negative emotions, especially disgust and contempt, and somewhat frequently surprise and interest emotions. These emotion clusters were then related to an overall satisfaction measure for the product. Results showed that the two most satisfied groups were the happy/content and the pleasant/surprise clusters. The happy/content cluster appeared slightly less satisfied than the pleasant/surprise group. Next in the level of satisfaction were the dissatisfied groups, which included the unemotional, unpleasant/surprise, and the angry/upset groups (in descending order).

As noted by Westbrook and Oliver (1991), their own work has three limitations that are addressed in the current research. First, their exploration of consumption-emotion patterns is limited to the categories of basic emotions developed in Izard's (1977) typology. Other typologies may yield different patterns of emotional response as well as different relationships to satisfaction. Second, the product category they used was automobiles. Other product categories may yield different patterns of emotional responses as well or a more generalized affect state can be realized. Lastly, the dimensionality of consumption-emotions and its relationship to satisfaction should be studied across multiple consumption contexts. Individuals may have different emotional responses depending on the nature of consumption behavior; that is, when consumption is for self or for other people.

PROPOSED RESEARCH

Our study extends the Westbrook and Oliver (1991) research by addressing these three limitations. Specifically, we examine the impact of emotions other than those provided by the Izard (1977) typology. The taxonomy provided by Izard (1977) and used by Westbrook and Oliver (1991) utilized a limited number of affect terms that included 7 negative emotions, 2 positive emotions and 1 neutral emotion. We include a larger number of positive emotions to provide a more diverse index of emotions.

Second, we examine consumer shopping behavior in general, instead of for a specific purchase. There is some evidence for a general affect state associated with consumer response to advertising (Aaker, Stayman and Hagerty 1986) and consumer satisfaction (Westbrook and Oliver 1991). Thus, by examining a diverse set of emotions, we may be able to isolate a core set of emotions experienced during shopping.

Third, we examine the emotions experienced while shopping for oneself and while shopping for others. Shopping affect may depend on the product being purchased as well as on the recipient. Dawson, Bloch and Ridgway (1990) posits that strong shopping motives are associated with positive emotional responses that occur in the marketplace. In addition, these emotional responses are particular to a specific context or situation. Thus, by exploring a shopper's satisfaction in terms of the recipient of the action (i.e., shopping for self or shopping for another), we hope to find differences in the affect experienced by these two shopping activities.

METHOD

Subjects

Two hundred forty-six undergraduate students participated in this study. One hundred twenty-seven students were recruited from introductory and social psychology courses at the University of Pittsburgh. The remaining one hundred nineteen participants were enrolled in a study skills course at Fisk University. The sample was composed of seventy-three (29.7%) males and one hundred seventy-three females (70.3%). One hundred twenty-six (51.2%) were black, one hundred thirteen (45.9%) were white, and 7 (2.8%) came from other minority groups. Our sample had an average age of 19.6 years old and 91.4% of subjects were between 17 and 22 years old.

Procedure

All participants completed a questionnaire that asked about themselves and various aspects of their shopping experiences. Students were asked to provide some descriptive data about themselves, including information about their education, future plans, and family background. Next, students were questioned about their shopping behavior.

An affect measure was employed in order to determine the kinds of emotions these shoppers experience during and after shopping. Specifically, participants were asked to rate the extent to which they experience each of eleven states while shopping for themselves and for others separately. Ratings were reported on a 1 (not at all) to 5 (very much) frequency scale. Students also completed the Westbrook and Black (1985) shopping motivation measure which involves considering how much satisfaction they derive from seventeen different types of shopping experiences, settings and outcomes. These items were rated on a 1 (none at all ) to 7 (a great deal) magnitude scale. After completing all measures, students were debriefed and thanked for their participation.

MEASURES

Affect States

The affect scale created in order to assess the types of emotions shoppers are likely to feel while shopping for themselves and for someone else was taken directly from Murrell, Frieze, Schmidt, Neo and Federouch (under review). The eleven emotions included were: happy, independent, silly, attractive, depressed, sad, guilty, confident, competent, helpful, and loving. These affect descriptors were adapted from Westbrook and Oliver (1991) with two changes. Several positive adjectives were added to include a range of affect experiences during shopping. And second, to be more applicable to general shopping activities, adjectives were selected that described general affect states. Participants used the affect scale to rate how they felt in two situations: while shopping for themselves and while shopping for others.

Shopping Motivation

Westbrook and Black's (1985) measure of shopping satisfaction consists of seven shopping motivation subscales. Anticipated utility, the first subscale, described the expectations of benefits or pleasure that the shopper hopes to receive by purchasing a product. Role enactment is a motivation which occurs when the consumer shops in order to fulfill a particular role which is related to shopping activity, such as being a housewife. Negotiation is shopping in order to bargain with salespeople over the price of the product. The fourth shopping motivation, choice optimization, involves searching for exactly the right product in the least amount of time. Shopping with the purpose of interacting with other individuals, either shoppers or salespeople, is called affiliation. The power and authority subscales refer to controlling or enjoying a position of higher status over someone else, usually a salesperson. Finally, stimulation motivation is at work when a shopper wants to see new and interesting surroundings. Westbrook and Black's reliability coefficient for these subscales were .64, .69, .54, .73, .67, .57, and .79 respectively. Most of the reliability coefficients from the present sample were higher than those reported than in the Westbrook and Black (1985) study, improving substantially for the anticipated utility and negotiation subscales, but decreasing slightly for power and authority subscale: .74, .73, .81, .73, .60, .72, .81. These differences could be explained by the sample differences in that college students typically have limited financial resources and would be more concerned about the utility and price merchandise and less over the use of power and authority, relative new dimensions for these individuals. The reliability coefficient for the overall shopping motivation measure was .84 for the present sample.

RESULTS

A k-means cluster similar to that used in Westbrook and Oliver (1991) was performed on the individual emotion ratings. The k-means clustering algorithm used indicated that a four-cluster solution of consumers produced both the most efficient results and interpretable solution. Five cluster solutions were examined to replicate that of Westbrook and Oliver (1990), with the four cluster solution being preferred, as it yielded the largest proportionate reduction in the trace of the within-clusters matrix, a measure of within-group homogeneity. These analyses were conducted separately for emotion states experienced while shopping for oneself and while shopping for others. In terms of emotions states experienced while shopping for oneself, four clusters were revealed. These clusters differed significantly across all emotion states (Wilk's Lambda = .067, F = 29.53, p<.0001 for self; Wilk's Lambda = .064, F = 29.56, p<.0001 for other).

We labeled Cluster 1 the positive/happy shoppers (n=20) were subjects who reported frequently on happy, independent, silly, attractive, confident, competent, helpful and loving emotions and infrequently on negative emotions. Cluster 2 was labeled the unemotional shoppers (n=31) reported infrequent experiences on all emotion states. Cluster 3 was labeled the negative/sad shoppers (n=60) and were subject to the highest frequency of depressed, sad, and guilty emotion states. Cluster 4 was labeled positive/efficacy (n=135) and was moderately high on positive emotions such as happy, independent, attractive and also high on instrumental emotions such as competent, confident, helpful and loving (see Figure 1).

For emotions experienced while shopping for others, we labeled cluster 1 (n=46) the negative/sad shoppers which was highest on the silly, depressed, sad, and guilty emotion states. Cluster 2 (n=82) was labeled the positive/average shoppers and was low across all emotion states. Cluster 3 (n=100) was labeled the positive/efficacy shoppers and was high on positive dimensions such as happy, independent, and attractive, and also on instrumental emotion dimensions such as confident, competent, helpful, and loving. Cluster 4 (n=18) represents the smallest cluster group and consistent of the unemotional shoppers. These individuals were the lowest across all emotion states (see Figure 2).

FIGURE 1

WHILE SHOPPING FOR SELF

Several similarities and a few differences emerged between the emotions clusters when shopping for oneself and when shopping for another person. In both cases the largest cluster group was the positive/efficacy group or those individuals who experienced both positive emotions (e.g., happy) and instrumental emotions (e.g., competent) when shopping for oneself or others. The second largest cluster group when shopping for oneself however, was the negative/sad affect cluster. These individuals experienced primarily negative emotions when shopping for themselves (e.g., sad). While this cluster was present for affect experienced when shopping for others, this group was small in number and also experienced feelings, labeled as "silly" as part of this negative affect. Interestingly, feeling silly when shopping for oneself was contained in the positive/happy cluster, suggesting that feeling silly represents a fun and frivolous type of emotion when shopping for oneself, but a foolish or nonsensical feeling when shopping for others.

To examine the relationship between patterns of consumer emotions and satisfaction, the mean satisfaction ratings were compared for each of the Westbrook and Black (1987) dimensions (utility, negotiation, power, affiliation, choice optimizing and stimulation) across the four cluster groups, examining self and other ratings separately. For emotions experienced while shopping for oneself, the cluster groups differed significantly across the utility dimension (F (3,242) = 3.79, p<.01), the negotiation dimension (F (3,241) = 5.86, p<.001), the affiliation dimension (F (3,242) = 5.42, p<.001) and the stimulation dimension (F (3,242) = 4.59, p<.004) of shopping satisfaction scale. The four clusters differed only marginally across the role dimension (F (3,239) = 2.19, p<.10) and were not significantly different for the choice optimizing dimension of shopping satisfaction. Post-hoc comparison (Tukey HSD tests) indicated that the positive/happy cluster was significantly higher in satisfaction on the utility, negotiation, affiliation, power and stimulation dimensions than the negative/sad cluster. In addition, the positive/happy cluster was significantly higher in satisfaction derived from the utility, affiliation, power and stimulation dimensions compared to the positive/efficacy cluster. This cluster was significantly higher than the unemotional cluster group on the utility and affiliation dimensions of consumer satisfaction (in all cases, p<.05). Thus, similar to the findings of Westbrook and Oliver (1991), the positive/happy cluster appears to have the highest level of satisfaction across the various satisfaction dimensions (see Table 1).

For emotions experienced while shopping for others, the four cluster groups differed significantly across the utility (F (3,242) = 4.80, p<.003), negotiation (F (3,242) = 2.74, p<.004), affiliation (F (3,242) = 3.95, p<.009), power (F (3,242) = 6.58, p<.001), and stimulation (F (3,242) = 4.04, p<.008) dimensions of consumer satisfaction. The four cluster groups did not differ across the role or choice optimizing dimensions of satisfaction. Post-hoc comparisons (Tukey HSD tests) revealed that the positive/efficacy cluster was significantly higher in satisfaction derived from the utility, negotiation, affiliation, power and stimulation dimensions than the unemotional cluster group. In addition, the positive/efficacy cluster was significantly higher in satisfaction derived from the utility dimension than the average cluster group and was higher in satisfaction derived from affiliation than the negative/sad cluster group (see Table 2).

FIGURE 2

WHILE SHOPPING FOR OTHERS

TABLE 1

MEANS FOR SATISFACTION EXPERIENCED WHILE SHOPPING FOR SELF EMOTION CLUSTERS

To determine the nature of the "core" emotion states experienced during shopping, subjects' responses to the eleven emotional states were factor analyzed using an oblique rotation for self and others' ratings, separately. Results revealed that when shopping for self, three factors emerged and accounted for 61% of the total variance. Factor 1, Efficacy (Eigenvalue=3.53) was composed of helpful, loving, confident, competent and attractive emotions. Factor 2, Negative (Eigenvalue=2.01), was composed of depressed, sad and guilty emotions. Factor 3, Positive (Eigenvalue=1.12), was composed of happy and independent emotions (see Table 3).

In terms of emotions experienced while shopping for others, two factors emerged and accounted for 61% of the total variance. Factor 1, Efficacy (Eigenvalue=4.16), contained confident, competent, independent, loving, happy, helpful and attractive emotion states. Factor 2, Negative (Eigenvalue=2.54), contained depressed, sad, guilty and silly emotion states (see Table 4).

TABLE 2

MEANS FOR SATISFACTION EXPERIENCED WHILE SHOPPING FOR OTHERS EMOTION CLUSTERS

TABLE 3

FACTOR ANALYSIS (SHOPPING FOR SELF)

DISCUSSION

The results of the present study showed that people tend to exhibit primarily positive/happy emotions (happy, independent, silly, attractive, confident, helpful and loving) while shopping for themselves and positive/efficacious emotions (happy, independent, attractive, confident, competent, helpful and loving) when shopping for other people. These results are consistent with the results of the Westbrook and Oliver (1991) study which showed that a large percentage of consumers experience positive emotions during consumption (e.g., 21% in the "happy/content" cluster and 23% in the pleasant- surprise cluster).

In addition, we found the positive/happy cluster was significantly higher on many of the satisfaction dimensions compared to other clusters when shopping for oneself. The positive/efficacy cluster was higher on many of the satisfaction dimensions than the other clusters while shopping for other people. These results show that people generate positive emotions when shopping, although they tend to be more goal-directed in their emotions while shopping for others than when they shop for themselves. Again, these results are consistent with that of Westbrook and Oliver (1991) who found that significantly higher levels of satisfaction among the happiness/contentment and pleasant-surprise patterns across all clusters. An implication of this finding is that emotion and satisfaction is contingent upon the context in which the activity takes place. One direction for future research is investigate whether contexts drive affect or whether context colors a consumer's affect judgments.

The core emotions identified in this research while shopping for oneself were efficacy, negative and positive. Efficacy and negative were the core emotions identified while shopping for others. These results showed that primarily positive emotions were exhibited while people shopped for themselves. When they shopped for others, the efficacious and positive emotions merged under one core emotion which we labeled efficacy. One implication of this finding is that when people shop for themselves, they may do so because of the pleasure of shopping, whereas when they shop for others, they tend to be goal-directed, a characteristic not unusual in the marketplace.

Another implication of this finding of core emotions is that it goes beyond simply being product-specific. Westbrook and Oliver (1991) had only looked at consumption of automobiles, a focus too narrow to capture the full effect of consumer evaluations. As stated by Westbrook (1980), being product-specific may not be particularly important to all consumers. Therefore, by looking a behavioral context such as shopping, core emotions yielded may be more generalizable to other consumer contexts.

Overall, satisfaction in shopping for self is higher than for other. This suggests that high levels of satisfaction and positive affect was higher for self because of an exposure effect. In other words, subjects shopped more frequently for self than other due to familiarity and exposure to the activity, thus increasing the affect and satisfaction experienced. A closer examination of emotions demonstrated by this study will help marketers better understand the role of emotions in consumer behavior.

TABLE 4

FACTOR ANALYSIS (SHOPPING FOR OTHERS)

A contribution of this paper is that it looks at satisfaction and emotional responses within a behavioral context, and not solely in a product-specific context. Our research has found that consumers tend to generate a more positive range of emotions when they shop, than previously observed in the literature. We have also shown that positive emotions have more dimensions than merely joy and interest as noted by Westbrook and Oliver (1991). We have incorporated the various dimensionalities of satisfaction and consumption emotions to provide a better understanding of the two constructs, and examined core emotions that are generated by consumers when they shop. One limitation of this research is the use of self-reporting rating scales to tap into the affective dimensions. However, we feel that these findings are nonetheless important because the link between emotion and satisfaction has implications for product evaluations, preferences, and choice.

REFERENCES

Aaker, D.A., Stayman, D.M. and Vezina R. (1988), "Identifying Feelings Elicited by Advertising," Psychology and Marketing, 5 (Spring), 1-16.

Aaker, D.A. and Hagerty, M.R. (1986), "Warmth in Advertising: Measurement, Impact, and Sequence Effects," Journal of Consumer Research, 12 (March), 365-381.

Clark, M. and Isen, A. M. (1982), "Toward Understanding the Relationship Between Feelings States and Social Behavior, " in Cognitive Social Psychology, eds. Albert Hastoff and Alice Isen, New York: Elsevier/North-Holland, 73-108.

Dawson, S., Bloch P.H., and Ridgway, N.M. (1990), "Shopping Motives, Emotional States, and Retail Outcomes," Journal of Retailing, 66 (Winter), 408-427.

Gardner, M. P. (1985), "Mood States and Consumer Behavior: A Critical Review," Journal of Consumer Research, 12 (December), 281-300.

Holbrook, M. B. and Batra, R. (1987), "Assessing the Role of Emotions as Mediators of Consumer Responses to Advertising," Journal of Consumer Research, 14 (December), 404-420.

Izard, C. E. (1977), Human Emotions, New York: Plenum.

MacInnis, D.J. and Park, C.W. (1991), "The Differential Role of Characteristics of Music on High-and Low-Involvement Consumers' Processing of Ads," Journal of Consumer Research, 18 (September), 161-173.

Murrell, A.J., Frieze, I.H., Schmidt, L.C., Neo, M., and Federouch, A.G. (under review), "The Impact of Shopping Motivation on Consumer Behavior and Affect: A look at Young Adult Shoppers," Journal of Retailing.

Murry, J.P. Jr., Lastovicka, J.L., and Singh, S.N. (1992), "Feeling and Liking Responses to Television Programs," Journal of Consumer Research, 18 (March), 441-451.

Olney, T.J., Holbrook, M.B., and Batra, R. (1991), "Consumer Responses to Advertising: The Effects of Ad Content, Emotions, and Attitude toward the Ad on Viewing Time," Journal of Consumer Research, 17 (March), 440-453.

Polivy, J. (1981), "On the Induction of Emotion in the Laboratory: Discrete Moods or Multiple Affect States?" Journal of Personality and Social Psychology, 41, 803-817.

Russell, J. A. (1979), "Affective Space Is Bipolar," Journal of Personality and Social Psychology, 37 (September), 345-356.

Schwartz, J. C. and Shaver, P. (1987), "Emotions and Emotion Knowledge in Interpersonal Relations," Advances in Personal Relationships, 1, 197-241.

Shaver, P., Schwartz, J., Kirson, D. and O'Connor, C. (1987), "Emotion Knowledge: Further Exploration of a Prototype Approach," Journal of Personality and Social Psychology, 52 (June), 1061-1086.

Westbrook, R. A. (1980), "Intrapersonal Affective Influences upon Consumer Satisfaction with Products," Journal of Consumer Research, 7 (June), 49-54.

Westbrook, R. A. and Black, W. C. (1985), "A Motivation-Based Shopper Typology," Journal of Retailing, 61, 78-103.

Westbrook, R. A. (1987), "Product/Consumption-Based Affective Responses and Postpurchase Processes," Journal of Marketing Research, 24 (August), 258-270.

Westbrook, R. A. and Oliver, R. L. (1991), "Dimensionality of Consumption Emotion Patterns and Consumer Satisfaction," Journal of Consumer Research, 18 (June), 84-91.

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