Advances in Consumer Research Volume 20, 1993      Pages 70-75

LINKING EMOTIONS AND VALUES IN CONSUMPTION EXPERIENCES: AN EXPLORATORY STUDY

Debra A. Laverie, Arizona State University
Robert E. Kleine III, Arizona State University
Susan Schultz Kleine, Arizona State University

ABSTRACT

This study examines the link between emotions and values in consumption experiences. A conceptual relationship between the two constructs is developed. The proposed relationship is tested in an exploratory study that elicited subject's own consumption experiences and then measured emotions and values on a paper and pencil instrument. The data support the premise that emotions and values are related in consumption experiences. Also, the data suggest that the self is the latent variable that links emotions and values in consumption experiences.

We all spend much of our lives consuming products. How does consumption make us feel? What are the values we obtain from, or express through, our consumption? And how might these feelings and values be linked through consumption experiences? People respond to and choose activities that make them feel a certain way (Havlena and Holbrook 1986) and correspond to important values (Munson 1984). Empirically, consumption experiences are linked with both values (Beatty, Kahle, Homer, and Misra 1985) and emotions (Havlena and Holbrook 1986; Havlena, Holbrook, and Lehman 1989). However, except for Holbrook's (1986) conceptualization of consumption experience, these two constructs have not been studied jointly in association with consumption.

Do consumers associate certain emotions with certain types of value in consumption experiences? For example, the emotions one associates with the playful or fun value of a consumption experience may be very different from the emotions one associates with a consumption experience that reflects the value of being well respected. The purpose of this project is to explore empirically the connection between patterns of emotions and values as they are linked through consumption. In the next section we discuss the emotion and values constructs and their potential linkage. Then we report the results of a study which explores patterns of emotions and values associated with consumption experiences.

THE LINK BETWEEN EMOTIONS AND VALUES IN CONSUMPTION EXPERIENCES

Emotion

Empirical research demonstrates that emotions are related to consumption behavior (Havlena and Holbrook 1986). The important role that emotions play in consumer's lives suggests that emotions can explain behavior in situations where other constructs, such as attitude, do not account for all or even a significant portion of the variability in behavior. For instance, Allen, Machleit, and Schultz Kleine (1992) showed that emotions supplement attitude in predicting consumption behavior. In addition, Westbrook and Oliver (1991) demonstrated that emotions help illuminate satisfaction (attitudinal) responses to consumption.

Emotions, which are feelings linked to a specific behavior (Gardner 1984), represent a richer and more complex realm of phenomena than does the attitude construct (Allen, et al. 1992; Holbrook 1986). Emotions are either positive, negative, or mixed in valence. Complex and/or conflicting emotions are richer than attitudes. Therefore, attitude may be too simple to represent the complexity of many consumption experiences (Allen, et al. 1992). There are several psychological frameworks that define and operationalize emotion (e.g., Izard 1977; Mehrabian and Russell 1974; Plutchik 1980). The present research uses Izard's framework which "assumes that separate and discrete emotions exist and that each has measurable, experiential, and motivational properties" (Izard 1972, p. 85). In his Differential Emotions Theory Izard (1972) conceptualizes ten fundamental emotions: joy, surprise, anger, disgust, contempt, shame, guilt, fear, interest, and sadness. Izard's typology has initiated research on emotions in a variety of consumer research contexts (e.g., Allen et al. 1988, 1992; Batra and Ray 1986; Westbrook 1987; Westbrook and Oliver 1991).

Values

Value is an abstract and complex construct that can provide underlying continuity to behavior (Pitts and Woodside 1984). Following Rokeach (1973, p. 25), we construe values as enduring beliefs that a specific mode of behavior or end-state is preferred over other alternatives. Therefore, values are a major influence on human behavior (Parsons and Shils 1951). The view that values guide behavior is evident in literature from psychology, sociology, and organizational behavior (Izard 1977; Rokeach 1973; Tolman 1951). In an interdisciplinary analysis of personal values, Clawson and Vinson (1978) suggest that:

Values may prove to be one of the most powerful explanations of, and influences on, consumer behavior. They can perhaps equal or surpass the contributions of other major constructs including attitudes, product attributes, degrees of deliberating, product classification, and life style (p. 396).

Therefore, the consumption of a product can express or fulfill a certain value.

Consumers' preferences for certain values are likely to be expressed through consumption. For example, certain products and activities may be preferred by a person who values excitement (e.g., fast cars, mountain biking, bungee cord jumping). On the other hand, a person placing security as very important would be likely to have a different set of preferred products and activities (e.g., an airbag in their car, going for long walks in the country, attending religious services). Researchers have suggested that we need a better understanding of the links between values and behaviors, and special consideration of how values interact with situations (Beatty et al. 1985).

To study consumption related values, Kahle (1983) modified the Rokeach Value Survey (RVS) into a smaller subset of values that were person oriented and generalizable across many activities. The List of Values (LOV) includes: sense of belonging, excitement, fun and enjoyment, warm relationships with others, self fulfillment, being well respected, a sense of accomplishment, security, and self respect. The LOV approach is a parsimonious way to measure values as compared to the RVS or Values and Lifestyles (VALS) approaches. In addition the LOV contains many items that people say influence their daily lives (Kahle 1986; Beatty et al. 1985).

Values and Emotions

There are several interesting parallels between emotions and values. Psychologists (Tolman 1951; Izard 1977) and practioners (e.g., Mowen 1988; Clawson and Vinson 1978) agree that both emotions and values play an extremely important role in behavior. Values are central beliefs expressed through specific behaviors (Rokeach 1973) and emotions are feelings linked to specific behaviors (Gardner 1985).

Values, according to Holbrook (1986), involve preference and thus are directly tied to the positive and negative affective opposition of emotions. The value expressed in a consumption experience is the result of the emotions that accompany the consumption experience. Accordingly, emotions and values are intertwined in consumption (Holbrook 1986). Thus, it makes sense that both values and emotions would be linked in consumer behaviors.

Study Objective

The objective of this research is to explore the relationship between values and emotions in consumption experiences by employing methods to represent the patterns of emotions that occur during consumption experiences that are associated with certain values. Specifically, are there discrete patterns of emotion that are linked to patterns of values in consumption experiences? Toward this end we conducted a study that examined the link between values and emotions in consumption experiences.

PRETEST

We conducted a pretest to explore the relationship between emotions and values in consumption experiences. Specifically, the pretest assessed a methodology and the measures for exploring the link between emotions and values. Using a paper and pencil instrument, subjects (n=71) were exposed to two scenarios that described consumption experiences. Each of nine scenarios expressed a particular LOV value. Subjects described a consumption experience that expressed a specific LOV value. Based on this self-reported experience subjects responded to measures of emotions and values.

Due to space limitations the empirical findings of the pretest are not reported. Results showed that seven of the ten emotions were significantly different across values (anger, contempt, and interest were not significant). However, the results appeared to be hampered by two methodological limitations. One, following Beatty et al.'s (1985) procedure, we asked subjects to rank the LOV values based on their importance to the described consumption experience. This ranking of values was difficult for the subjects because values are so closely held (Munson and McIntyre 1979) and some subjects didn't follow directions carefully. In addition, rank order data precluded the use of many types of statistical analyses. Two, the manipulation check suggested that some subjects seemed to focus on the activity depicted in the scenario more than the expressed value. Procedures were altered for the main study to overcome these methodological limitations.

AN EXPLORATORY STUDY

We conducted a study to investigate the proposed link between emotions and values in consumption experiences. Value expressive consumption experiences were elicited from subjects after which they responded to emotion and values measures. Described next are the methods and subjects used in this study. We then report the results.

Eliciting Value Expressive Consumption Experiences

A consumption experience was defined for subjects as any activity they do while using a product. Several examples of consumption experiences were provided (e.g., eating, driving a car, wearing clothes, playing volleyball, and listening to music). Subjects were then asked to recall and describe a consumption experience that was important to them because of one of the nine LOV values. For the "warm relationship with others" value, for example, subjects were asked to think of an experience when they felt warm and happy because they were with good friends and family (how each value was defined for the subjects is described in Appendix 1).

Measures

Emotions. We used Izard's (1977) DES-II to measure subject's emotions during consumption experiences. The DES-II has been an effective measure of emotions for consumer researchers (e.g., Allen et al. 1992, 1988; Westbrook and Oliver 1991). The DES-II instructions ask the subjects to consider the experience they described and to rate how often s/he experienced each emotion item during the experience. Subjects rate 30 items, 3 for each of the ten fundamental emotions. Each item is administered on a 5-point never to very often scale.

Values. Kahle's (1983) List of Values (LOV) measured the values subjects associated with the consumption experiences they described. Based on the analytical limitations of rank ordering values revealed in the pretest the LOV values were measured using Likert scales. Values assessed using Likert scales are as reliable as those measured with the ranking procedure (Munson and McIntyre 1979).

Subjects were instructed to think about the experience they described and then to use the scales to indicate the importance of each value to their experience. Each value was measured on a 5-point extremely unimportant to extremely important scale. Subjects also rank ordered the nine LOV values to afford a manipulation check.

Procedures

Subjects were 131 male and female undergraduate students from five sections of a marketing course. Each subject independently described and answered questions about two consumption experiences. Due to incomplete data one response was eliminated leaving 261 usable consumption experiences. The values described in the consumption experience elicitation instructions were randomized across subjects. The cover sheet indicated that participation in the study was voluntary, the next sheet stated the study's purpose and presented subject instructions.

Subjects read a definition of a consumption experience and then read a description of one the nine LOV values. Subjects were asked to write out a description of a consumption experience that was important to them because of the value that they were exposed to. Finally, subjects responded to Izard's DES-II emotion measures and the LOV values on Likert scales. In addition, subjects ranked the LOV values according to their importance in the experience they had described.

RESULTS

The emotion and value variables were first factor analyzed to assess the dimensionality of each construct. The variables that resulted from the factor analyses were then analyzed via canonical correlation analysis to assess the emotion-value relationship.

Structure of the Emotion Measure

The a priori measurement model which specified 10 factors, each consisting of three items, was analyzed via confirmatory factor analysis (SAS's Proc CALIS). The overall fit of the ten factor model, analyzing the covariance matrix, was acceptable (c²(381)=620.66, p>.0001, Bentler and Bonett's Normed Index=.975, RMR=.213). All parameter estimates were reasonable and in the anticipated direction. The average variance extracted from the items was .59 (Bagozzi and Yi 1988). The distribution of the residuals is approximately normal. Only three residuals exceeded 2.0. Furthermore, the fit is comparable with other reported confirmatory factor analyses of Izard's measurement model (e.g., Allen, et al. 1988). Sum scales were constructed from the thirty items for the ten emotions as defined Izard (1977).

TABLE 1

EXPLORATORY FACTOR ANALYSIS OF VALUE ITEMS (VARIMAX ROTATED)

Structure of the Value Measure

The LOV values are suggested to be nine different values that are important in people's daily lives (Kahle 1983). In addition the LOV measure has demonstrated convergent and discriminant validity in past research (Beatty et al. 1985; Kahle 1983). This implies theCadmittedly strongChypothesis that if the LOV is submitted to a factor analysis nine factors will result.

An exploratory factor analysis of the scaled data on the nine LOV values using varimax rotation reveals three distinct factors (see Table 1). The first factor is defined by four highly loading values that all relate to interactions with others (e.g., a sense of belonging, warm relationships with others, security, and being well respected). Accordingly, we labeled this factor 'others'. The second factor is formed by two self related outcomes and is labeled 'self' (e.g., sense of accomplishment and self fulfillment). The value self-respect cross loads on the 'others' factor which intuitively makes sense since self respect is often governed by one's perception of what other people think about him or her. Due to the heavy cross loading this item is not used in further analyses. The third factor is labeled 'hedonic' and is made up of the values of fun/enjoyment and excitement.

These results suggest that the LOV consists of three factors in the context of consumption experiences. Kennedy, Best, and Kahle (1988) report a similar factor structure. Based on the results of the factor analysis three sum scales were constructed for the value items that loaded together on a factor. These sum scales are used in the canonical correlation analysis reported below.

Elicitation Assessment

To check the efficacy of instructions we compared the top-ranked value to the value the subjects's consumption experience was supposed to express. On average 76% of the subjects ranked the value described in their consumption experience as the most important (e.g., number one). The percentage of subjects who ranked the value described in their stimulus as number one ranged from 95% for a 'sense of accomplishment' to 57% for 'well respected'. This suggests that the nine values are well represented by subject's consumption experiences.

The Relationship Between Values and Emotions

We explored the relationship between emotions and values in consumption experiences using canonical correlation analysis. Specifically, the three values factors (e.g., others, self, and hedonic) were related to the ten emotions (e.g., joy, surprise, anger, disgust, contempt, shame, guilt, fear, interest, and sadness). Three significant canonical variates were obtained (see Table 2).

The loadings on the first canonical variate suggest that consumption experiences expressing hedonic values (e.g., fun and enjoyment and excitement) are associated with the emotions interest, surprise, and enjoyment. This variate accounts for much (37%) of the shared variance between emotions and values.

The second canonical variate suggests that interest and fear are the most salient emotions in consumption experiences that express self related values (e.g., sense of accomplishment and self fulfillment). This link accounted for 29% of the shared variance.

The third variate suggests a relationship between the others related value and the emotions of shame and decreasing surprise in consumption experiences. However, this variate accounts for only 9% of the shared variance and therefore should be interpreted with caution, despite its statistical significance (Pedhazur 1982).

Anger, sadness, disgust, contempt, and guilt, all negative emotions, did not help in explaining the linear combinations between emotions and values. The emotions related to values are typically positive. In addition, when people are asked to recall something related to a value and write about it they typically pick a positive experience to discuss (Schrum, McCarty, and Loeffler 1990).

TABLE 2

CANONICAL CORRELATION ANALYSIS BETWEEN VALUES AND EMOTIONS

DISCUSSION

The results offer encouraging support for the proposed link between certain emotions and values in consumption experiences. These findings are a step toward a better understanding of post-purchase consumer behavior. At a minimum, these results encourage further examination of the link between values and emotions as facets of consumption experiences. In addition, the results suggest some methodological considerations that need to be addressed in future research.

The Relationship Between Values and Emotions

The results of the canonical analysis suggest that certain emotions are linked to consumption experiences that express different values. Consumption experiences that are valued because they are enjoyable are linked to positive emotions. On the other hand, experiences that are valued because of their link to the self are combined with interest and a negative emotion, fear. Values that center around others are linked to shame and decreasing surprise. It is interesting to note the parallels between this exploratory study's findings and the literature on the self.

The social-psychological literature suggests three facets of the self that can be linked to the three value variables. The 'hedonic', 'self', and 'others' values parallel the hedonic, private, and public aspects of the self (e.g., Greenwald and Breckler 1985). The 'hedonic' value relates to three emotions that all involve arousal (interest, surprise, and enjoyment). The results suggest that this variate represents a hedonic portion of the self. This aspect of the self is hedonically guided toward positive affective states (Greenwald and Breckler 1985). In addition, the hedonic self is a condition of not distinguishing sharply between self and others (Greenwald and Breckler 1985; Schlenker 1985). This state is evident in the first canonical variate. The 'others' and 'self' values load heavily and equally on this 'hedonic' variate.

The second variate labeled 'self' resembles the private self. The private self allows for self-evaluation in the absence of others (Greenwald and Breckler 1985). This condition is clear in the canonical loadings, the 'others' value loads negatively on the 'self' variate. A major task of the private self is individual achievement. The value items that load on the 'self' factor (a sense of accomplishment and self-fulfillment; see Table 1) evidence this.

The emotions related to the private self are interest and fear. Intuitively one would expect interest to be important to the private self as this aspect of the self is guided by internal standards (McClelland, Atkinson, Clark, and Lowell 1953). On the other hand, it may seem surprising that fear is an important emotion in the private self. However, the private self is developed based on the internalization of the evaluative standards of significant others (Greenwald and Breckler 1985). For example, a respondent wrote: "I felt excited and scared going through sorority rush. I was terrified because I .Ê.Ê. was not sure if I would fit in. .Ê.Ê. I was scared, but I hoped that what I was wearing would make me feel comfortable and help me to fit in." Therefore, the emotion of fear may be linked to the concern of meeting the standards one has adopted for him/herself.

The third variate depicts the public self. The public self is sensitive to the evaluations of others and seeks to win the approval of those who are important to the individual (Greenwald and Breckler 1985). An important task of the public self is earning credit in exchange relationships with others. These characteristics are evident in the values that form the 'others' factor, namely 'being well-respected', a 'sense of belonging', 'warm relationships with others', and 'security'.

The emotions that are linked to the public self are shame and decreasing surprise. Although it may seem surprising that shame is an important emotion in the public self, shame is the primary social emotion (Goffman 1967; Lewis 1971; Scheff 1990; Rosenberg 1979). Shame is an emotion that arises from the monitoring of one's own actions by viewing one's self from the standpoint of others (Scheff 1990). Thus, it makes sense that shame is an important emotion in the public self which is concerned with what other people think of us. Likewise, the public self is adverse to the potential embarrassment of being caught by surprise in social situations. In sum, the results of the canonical correlation analysis suggest some interesting relationships between emotions and values in consumption experiences. It is encouraging that these relationships can be explained by the extensive literature on the self. Furthermore, these results lead us to speculate that the self is the underlying link between values and emotions in consumption experiences.

Limitations and Future Research

The specific findings of this study cannot be generalized without caution. As noted earlier, the research is exploratory in nature. In addition, both the values and the emotions measures are based on retrospection. Future research measuring emotions and values during the consumption experience would provide important insight into consumption experiences.

Measurement of Emotions. Most measures of emotions, such as Izard's scale used in this research, are essentially based on aided recall (Westbrook 1987). This research enhanced subjects' recollection by asking them to write a detailed description of their experience. However, alternative ways of tapping emotions may yield different information about consumption experiences. For instance, measurement directly after an experience would likely lead to richer information as compared to the emotions one can recall. As noted in the findings of this research typically just a few emotions are associated with certain values. Perhaps this is because the most important one or two emotions are all that subjects recall and they may have experienced a much larger range of emotions during the experience. Identification of the best approach for predicting consumption behavior is a task for future research.

Another question consumer researchers need to address is the appropriateness of psychological measurement schemes for consumption activities. The findings of this research suggest that the many negative emotions in Izard's scale may not be suited to value expressive consumption experiences since values are typically positive. It appears that consumer researchers should explore alternative measurement frameworks or develop a measurement procedure that is more appropriate for consumption behavior. The qualitative measurement of emotions (Smith-Lovin 1990) is one promising alternative.

Examining the Dimensions of Values. This research suggests that the dimensionality of Kahle's (1983) LOVs needs to be assessed further in the context of consumption experiences. Although designed to reflect values that consumers experience in their daily lives (Beatty et al. 1986), our findings suggest that the LOV values represent three distinct factors. This encourages further examination of the validity of this construct in the context of consumption experiences. The value factors offer a parsimonious view of values in the consumption context.

In addition, the study supports measuring values on Likert type scales. This produced better quality data and more flexibility in analysis. The Likert type approach does not force the respondent to rank order values that may be equally important and it allows for differences in the intensity with which a particular value is held (Munson 1984; Munson and McIntrye 1979; Clawson and Vinson 1973).

Emotions and Values in Consumption Experiences. Both emotions and values are rich constructs for understanding post-purchase consumer behavior. The results of this study provide evidence of a connection between patterns of emotions and values in consumption experiences. Perhaps consumers choose certain consumption experiences because of this linkage. Researchers whose interest is pre-purchase phenomena might explore how the effectiveness of promotional messages that reflect specific values and emotions differ across common segmentation variables. Furthermore, different aspects of the self may help explain the emotion-value linkage. Future research should address this role of the self in consumption experiences.

APPENDIX

The following was used to prompt a consumption experience that expresses a particular LOV value: "Think of an experience when you were using products and ________." One of the following phrases completed the sentence: 1) you were well respected; 2) you were self-fulfilled; 3) you felt a sense of accomplishment; 4) you felt a sense of security; 5) you felt a sense of belonging; 6) you felt self respect; 7) you were excited; 8) you felt warm relationships with other people; 9) you had fun.

REFERENCES

Allen, Chris T., Karen A. Machleit, and Susan S. Marine (1988), "Assessing the Emotionality of Advertising Via Izard's Differential Emotions Scale," in Advances in Consumer Research, Vol.15, ed. Michael J. Houston, Provo, UT: Association for Consumer Research, 226-231.

Allen, Chris T., Karen A. Machleit and Susan Schultz Kleine (1992), "A Comparison of Attitudes and Emotions as Predictors of Behavior at Diverse Levels of Behavioral Experience," Journal of Consumer Research, 18 (March), 493-504.

Bagozzi, Richard P. Youjae Yi (1988), "On the Evaluation of Structural Equation Models), Journal of the Academy of marketing Science, 16 (1), 74-93.

Batra, Rajeev and Michael Ray (1986), "Affective Responses Mediating the Acceptance of Advertising," Journal of Consumer Research, 13 (September), 234-249.

Beatty, Sharon E., Lynn R. Kahle, Pamela Homer, and Shekhar Misra (1985), "Alternative Measurement Approaches to Consumer Values: The List of Values and Rokeach Value Survey," Psychology and Marketing, 2 (Number 3), 181-200.

Clawson, C.J. and Donald E. Vinson (1978), "Human Values: An Historical and Interdisciplinary Analysis," in Contributions to Consumer Research V, ed. H. Keith Hunt, Chicago: Association for Consumer Research, 396-402.

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

Goffman, Erving (1967), Interaction Ritual, New York: Anchor.

Greenwald, Anthony G. and Steven J. Breckler (1985), "To Whom is the Self Presented?", in The Self and Social Life, ed. Barry R. Schlenker, New York: McGraw-Hill.

Havlena, William J. and Morris B. Holbrook (1986), "The Varieties of Consumption Experience: Comparing Two Typologies of Emotion in Consumer Behavior," Journal of Consumer Research, 13 (December), 394-404.

Havlena, William J., Morris B. Holbrook and Donald R. Lehmann (1989), "Assessing the Validity of Emotional Typologies," Psychology and Marketing, 6 (Summer), 97-112.

Holbrook, Morris B. (1986), "Emotion in the Consumption Experience: Toward A New Model of Consumer Behavior," in The Role of Affect in Consumer Behavior, eds. Robert A. Peterson, Wayne D. Hoyer, and William R. Wilson, Lexington, MA: Lexington Books, 17-52.

Izard, Carroll E. (1972), Patterns of Emotions, New York: Academic Press.

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

Kahle, Lynn R. (1983), Attitudes and Social Adaptation: A Person-Situation Interaction Approach, London: Pergamon.

Kahle, Lynn R., Sharon E. Beatty, and Pamela Homer (1986), "Alternative Measurement Approaches to Consumer Values: The List of Values (LOV) and Values and Life Style (VALS), Journal of Consumer Research, 13 (December), 405-409.

Lewis, Helen B. (1971), Shame and Guilt in Neurosis, New York: International Universities Press.

Kennedy, Patricia F., Roger J. Best, and Lynn R. Kahle (1988), "An Alternative Method for Measuring Value-Based Segmentation and Advertisement Positioning," in Current Issues in Research in Advertising, Vol. 11, 139-155.

McClelland, D.C., Atkinson, J.W., Clark, R.A. and E.L. Lowell (1953), The Achievement Motive, New York: Appleton-Century-Croft.

Mehrabian, Albert and James Russell (1974), An Approach to Environmental Psychology, Cambridge, MA: MIT Press.

Mowen, John C. (1988), "Beyond Consumer Decision Making," Journal of Consumer Marketing, 5 (Winter), 15-25.

Munson, J. Michael (1984), "Personal Values: Considerations on Their Measurement and Application to Five Areas of Research Inquiry," in Personal Values and Consumer Psychology, eds. Robert E. Pitts, Jr. and Arch G. Woodside, Lexington, MA: Lexington Books, 13-33.

Munson, J. Michael and Shelby H. McIntrye (1979), "Developing Practical Procedures for the Measurement of Personal Values in Cross-Cultural Marketing, Journal of Marketing Research, 16 (February), 48-52.

Parsons, Talcott and Edward R. Shils (1951), Toward a General Theory of Action, Cambridge: Harvard University Press.

Pedhazur, Elazar J. (1982), Multiple Regression in Behavioral Research, New York: CBS College Publishing.

Pitts, Robert Jr. and Arch G. Woodside (1984), "Personal Values and Market Segmentation: Applying the Value Construct," in Personal Values and Consumer Psychology, eds. Robert E. Pitts, Jr. and Arch G. Woodside, Lexington, MA: Lexington Books, 55-67.

Plutchik, Robert (1980), Emotion: A Psychoevoluntionary Synthesis, New York: Harper and Row.

Rokeach, Milton (1973), The Nature of Human Values, New York: Free Press.

Rosenberg, Morris (1979), Conceiving the Self, New York: Basic Books.

Scheff, Thomas J., (1990) "Socialization of Emotions: Pride and Shame as Causal Agents," in Research Agendas in the Sociology of Emotions, ed. T. D. Kemper, Albany, New York: State University of New York Press.

Schrum, L. J., John A. McCarty, and Tamara L. Loeffler (1990), "Individual Differences in Value Stability: Are We Really Tapping True Values?," in Association for Consumer Research, Vol. 17, eds. Gerald J. Gorn and Richard W. Pollay, Provo, UT: Association for Consumer Research, 609-615.

Smith-Lovin, Lynn (1990), "Emotion and the Confirmation and Disconfirmation of Identity: An Affect Control Model," in Research Agendas in the Sociology of Emotions, ed. Theodore D. Kemper, Albany, New York: State University of New York Press, 238-270.

Tolman, Edward C. (1951), "A Psychological Model," in Toward A General Theory of Reasoned Action, eds. Talcott Parsons and Edward A. Shils, Cambridge: Harvard University Press, 148-163.

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

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

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