The Effects of Context-Induced Mood States on Initial and Repeat Product Evaluations: a Preliminary Investigation

ABSTRACT - This study examines the effects of context-induced mood states on initial and repeat global product evaluations, attitude toward the object, and purchase intentions. The on-line and memory-based processing paradigms are used to predict and interpret the effects of positive and negative mood states on product evaluations, and purchase intentions measured at two different times. Consistent with prior findings, the results indicate that subjective affective states influenced initial product evaluations. In contrast, repeat product evaluations were not susceptible to mood states. Implications for advertising, media planning, and retailing are discussed and further research suggestions are offered.

Citation:

John Hadjimarcou, John W. Barnes, and Richard S. Jacobs (1996) ,"The Effects of Context-Induced Mood States on Initial and Repeat Product Evaluations: a Preliminary Investigation", in NA - Advances in Consumer Research Volume 23, eds. Kim P. Corfman and John G. Lynch Jr., Provo, UT : Association for Consumer Research, Pages: 337-341.

Advances in Consumer Research Volume 23, 1996      Pages 337-341

THE EFFECTS OF CONTEXT-INDUCED MOOD STATES ON INITIAL AND REPEAT PRODUCT EVALUATIONS: A PRELIMINARY INVESTIGATION

John Hadjimarcou, University of TexasBEl Paso

John W. Barnes, University of TexasBEl Paso

Richard S. Jacobs, University of TexasBEl Paso

ABSTRACT -

This study examines the effects of context-induced mood states on initial and repeat global product evaluations, attitude toward the object, and purchase intentions. The on-line and memory-based processing paradigms are used to predict and interpret the effects of positive and negative mood states on product evaluations, and purchase intentions measured at two different times. Consistent with prior findings, the results indicate that subjective affective states influenced initial product evaluations. In contrast, repeat product evaluations were not susceptible to mood states. Implications for advertising, media planning, and retailing are discussed and further research suggestions are offered.

INTRODUCTION

Findings in marketing and social psychology provide strong evidence for the existence of mood effects on various aspects of behavior (e.g., Isen 1970), cognition (e.g., Bower 1981; Bower, Gilligan, and Monteiro 1981; Knowles, Grove, and Burroughs 1993), and judgments of familiar and unfamiliar stimuli (e.g., Johnson and Tversky 1983; Gardner 1985; Batra and Stayman 1990). In general, people in positive moods have strong tendencies to provide positive evaluations and act in positive ways, while people in negative moods tend to do the reverse (Clark and Isen 1982; Gardner 1985). However, the study of the impact of mood states on judgment and behavior has thus far focused on the salience of mood states on initial product judgments, while their influence on repeat judgments remains largely ignored in the literature. Further, in most situations consumers receive product information and may form initial evaluations of a product in the context of a particular mood state (for example, forming an initial judgment of a product advertised in a commercial during a happy or sad TV program). At a later time, they may be called upon to reconsider the advertised product in an actual purchasing encounter in the context of yet another mood state. An intriguing question is whether mood states are salient at both stages of this purchasing scenario. Although advertisers keep mood contexts in mind when developing advertising strategies and media plans (Goldberg and Gorn 1987), a limited understanding of mood effects on repeat product evaluations and purchase intentions may result in inappropriate advertising strategies with costly implications. Greater understanding of such effects is thus needed by researchers and practitioners alike.

The purpose of this study is to investigate the influence of subjective affective states on initial as well as repeat product evaluations. Specifically, we will examine the case where subjects in a positive or negative mood receive information about an ostensibly unfamiliar stimulus which they are asked to evaluate immediately following the presentation of information, as well as after a two-day interval. We also argue that the paradigms of on-line and memory-based processing could be used to accurately predict the direction of both initial and repeat measurements involving global product evaluations, attitude toward the object, and purchase intentions. First, we will present a review of the relevant literature and the conceptual foundations of our study. Following the development of the hypotheses, we describe the study methods, processes, and analysis of the data. Finally, we report the results and briefly discuss implications for future research.

PAST RESEARCH AND CONCEPTUAL FOUNDATIONS

Considerable evidence exists to suggest that mood states affect judgments, behavior, and the recall of information in a mood congruent direction. In particular, using mood congruency theory, researchers agree that subjects processing information in a positive mood state rate ambiguous stimuli as more pleasant (Isen and Shalker 1982), concentrate on positive rather than negative self-relevant information (Mischel, Ebbesen, and Zeiss 1973), and reward themselves more generously (Mischel, Coates, and Raskoff 1968) than those in either a neutral or negative mood state. Further, Bower, Gilligan and Monteiro (1981) examined the impact of mood on cognitions and concluded that subjects were more likely to recall information congruent with their mood at encoding. Bless, Mackie, and Schwarz (1992) also reported that subjects had the tendency to reduce cognitive effort while in a positive mood.

In a marketing context, studies have found that subjects processing information in a positive mood are more likely to provide more positive brand attitudes (e.g., Batra and Stayman 1990). Positive mood states have also been linked to higher levels of persuasion, especially with regard to advertising claims (Batra and Stayman 1990). Similarly, Goldberg and Gorn (1987) reported that viewers who watched commercials during a happy television program were more likely to provide positive cognitive responses and perceive the commercials as more effective than those who watched the commercials in the context of a sad program.

While the congruent effects of positive moods have been substantiated in the literature, the influences of negative mood states have been less predictable (Clark and Isen 1982; Gardner 1985). For example, some researchers have found that being in a negative mood state increases one's antisocial behavior (Moore, Underwood and Rosenhan 1973). At the same time, others have reported that some negative feelings increase prosocial behavior (see Clark and Isen (1982) for an excellent review of the various findings).

In summary, past research focusing on mood states and consumer behavior has reinforced the notion that consumer product evaluations and recall of ad and/or product related information are significantly affected by the consumer's subjective mood. In many cases, mood congruency theory accurately predicts the impact of positive and negative mood states upon salient consumer behavior variables. However, past research has concentrated solely on the impact of mood states on initial product judgments, and thus ignored more externally valid marketing situations that include repeat evaluations. Recent studies have further shown that mood states may not have an impact on memory-based evaluations as compared to on-line evaluations (Srull 1987; Knowles, Grove, and Burroughs 1993). On-line evaluations involve a simultaneous consideration of available information about a stimulus with "...an implicit or explicit objective of making an evaluation..." (Knowles, Grove, and Burroughs 1993, p. 136). Evaluations via memory-based processing involve the retrieval of information already stored in long-term memory.

Srull (1987), for example, observed congruent mood effects on initial evaluations taken 48 hours after product information was presented. In this case, subjects were initially asked to form an evaluation of the product (on-line condition) while in a mood state. In contrast, later (but initial) evaluations of subjects first instructed to simply "comprehend" the product information (memory-based condition) failed to exhibit a mood-congruent pattern, even though the subjects also experienced different mood states at encoding (Experiment 1). In the same study, the author investigated the impact of mood states on later evaluations when moods were induced at the time product evaluations were measured (at retrieval). In this case, the results suggested that mood states may affect memory-based evaluations provided they are present when evaluations actually occur (Experiment 2). In the same experiment, Srull (1987) further reported that subjects initially instructed to use the information to form a product evaluation (on-line condition) did not exhibit different evaluations, counter to the mood congruency theory. Consistent with these findings, Knowles, Grove, and Burroughs (1993) reported that delayed mood states did not influence brand evaluations in a memory-based condition, but did affect the amount and type of information recalled by the respondents. Neither of these studies, however, has examined the salience of mood states on repeat product evaluations since the measurements were taken only once (either at encodingCTime One or retrievalCTime Two).

Yet, the majority of consumption situations involve a two-stage process (Keller 1991; Knowles, Grove, and Burroughs 1993). Consumers receive information about products via print and electronic media and may form initial product evaluations at that time. Often the information becomes available while watching a mood-inducing happy/sad TV program, or after reading a happy/sad article in a popular magazine. Consumers may later be called upon to re-evaluate the product in an actual purchase encounter (e.g., in a retail establishment) while they are experiencing a similar or a different mood state. It would be interesting then to examine the influence of mood states at both stages of this consumption situation.

On-line evaluations are quite common, whereas memory-based evaluations are relatively infrequent (Hastie and Park 1986). The authors suggest that most individuals form initial predispositions toward a stimulus spontaneously when receiving information about it. This occurs regardless of specific instructions designed to prevent the formation of evaluations. Moreover, simulation of a memory-based situation in an experimental setting may prove to be difficult in itself (Hastie and Park 1986). Nonetheless, it may be impossible to determine whether the subjects formed an overall evaluation of the product at encoding (initial evaluation) or retrieval (perhaps, a repeat evaluation). In this study, therefore, subjects were instructed from the outset to form overall evaluations of the stimulus while stimulus-related information was immediately available to them (on-line evaluation). The subjects were then instructed to provide repeat (not memory-based) evaluations of the stimulus at a later time. Unlike previous research, it is interesting to note that different affective states were induced at both evaluation stages.

HYPOTHESES

Given the previous discussion on the mood congruency effects, we would expect initial evaluations to be consistent with previous findings. Therefore, we developed the following hypothesis:

H_1: Subjects in a positive context-induced mood will provide more favorable global evaluations, attitude toward the object, and purchase intentions than subjects in a negative mood state.

Repeat evaluations cannot readily be considered memory-based evaluations since subjects have already formed their overall product evaluations at an earlier stage. To reiterate, memory-based evaluations are those occurring at a subsequent time, while information for forming these evaluations was received at an earlier time. The basic premise here is that subjects have not supposedly formed an evaluation at the time stimulus information was received. Therefore, it is expected that when subjects are later called upon to make an evaluation they will retrieve the previously stored information for that purpose. In the present study, subjects were specifically instructed to form evaluations on line. Thus, the impact of mood states on repeat evaluations should be minimal (cf. Srull 1987):

H₂: Subjects in either a positive or negative context-dependent mood state will provide repeat global evaluations, attitude toward the object, and purchase intentions that are similar to initial evaluations.

METHODOLOGY

Design and Procedure

The subjects were 81 undergraduate students enrolled in marketing courses at a major Southwestern university. They were asked to participate in the study on a voluntary basis, and received extra credit in their class for participation. The experiment was conducted during regularly scheduled class periods. The respondents were also told that they would be participating in two unrelated studies conducted over two consecutive class meetings (or stages).

The subjects were informed that their Stage 1 participation would involve two different studies: an "Empathy Study" and a "Product Evaluation Study." Two days later, the subjects were asked to participate in Stage 2 of the experiment. At that time, they were told that "...preliminary examination of the data collected two days ago led us to believe that we need to collect some additional information..." Otherwise, the procedures for Stage 2 were identical to the procedures used in Stage 1.

Consistent with Batra and Stayman (1990) and Gardner (1992), the experiment was conducted using two separate research assistants in an attempt to mask the connection between the two studies. In addition, the two separate survey instruments were printed on different colored paper, using two different font formats. Later debriefing of the subjects revealed that the participants saw no connection between the two studies either in Stage 1 or Stage 2.

Phase I. As mentioned above, each stage of the experiment included two phases. The first phase of Stage 1 involved the manipulation of context induced-mood states. This was coined as the "Empathy Study." During this phase, subjects were randomly assigned to one of two experimental groups. After completing an informed consent form, forty-one subjects were asked to "...take a few minutes to recall a happy event in your life ...and to write down exactly how you felt during that period." This was done to prime the subjects' affective states prior to reading a similarly-valenced (i.e., happy) story. In contrast, forty subjects were asked to recall a sad event in their life as well as to elaborate on their feelings at the time prior to reading a "sad" story.

In the first phase of Stage 2, approximately half of the subjects that read the happy story in Stage 1 were asked to read a happy story, and the remainder to read a sad story. Similarly, half of the subjects that read the sad story in Stage 1 were asked to read a happy story and vice versa. As in Stage 1, assignment to the two mood treatments in Stage 2 was random.

All four stories (two happy, two sad) were selected from a total of 12 stories pretested at an earlier time using 96 subjects similar to the ones participating in this study. The stories were generated from actual stories found in popular magazines such as Reader's Digest. After reading the stories, pretest subjects were asked to respond to the Mood Short Form (MSF) scale (Peterson and Sauber 1983) to assess their mood state. The four stories ultimately chosen were those eliciting the most positive and negative mood states in our pretest subjects. For example, one of the happy stories described how a middle-aged doctor saved the life of his long-lost early childhood friend, suffering from leukemia, without either of the two realizing at first their early childhood ties. The story becomes happy when both become reacquainted. Another story, designed to elicit a negative mood, featured a pregnant woman's battle with cancer. She ultimately gives birth to a healthy baby boy, despite the massive chemotherapy treatments received during her pregnancy. Soon after the birth, she also finds out that her cancer is in full remission. Tragically, both her husband and baby boy are then involved in a deadly automobile accident.

After reading the story, the subjects were asked to record their feelings and thoughts. Consistent with Gardner (1992), this was done to accentuate the story's impact on mood states. The subjects were then asked to respond to the MSF scale.

Phase II. The second phase of the experiment was conducted immediately following completion of the "Empathy Study." Another researcher asked the participants to complete a second informed consent form and briefly explained the purpose of the study. The subjects were first asked to read information about a new digital audio tape player (code named DAT-111) soon to be introduced in the market. Specifically, the subjects were instructed to "... imagine that you are actually in the market to buy the DAT player ... try to get an idea of what it would be like and of its quality ... try to form as clear an impression of the product as you can. Later, we will be asking you to evaluate the DAT player and its features, and also whether you would purchase it..." The instructions were specifically designed to simulate an on-line processing situation (cf. Hastie and Park 1986; Srull 1987; Knowles, Grove and Burroughs 1993). The information consisted of a total of 15 attributes describing various features of the DAT (e.g., it comes with many accessories, average sound quality of 6.7 on a scale of 1 to 10). The attribute information was arranged in random order and presented in a format similar to that found in Consumer Reports.

All 15 attributes were derived from a total of 30 attributes pretested at an earlier time. Specifically, 96 undergraduate students from the same population in which the main study was conducted were asked to evaluate the 30 attributes in terms of favorableness on a seven-point Likert-type scale (1-Very unfavorable to 7-Very favorable), and importance in a purchase decision (1-Very unimportant to 7-Very important). Based on these results, the attribute information selected was evaluated as relatively neutral in terms of its favorability. In addition, information was chosen that was considered more or less equally important to the subject population. This was done to avoid biasing for or against any discrete attribute information. We wanted each piece of attribute information to be given an approximately equal chance of being considered in product evaluation regardless of mood state. Consequently, the product could not be clearly evaluated as either positive or negative, but was otherwise regarded as ambiguous in terms of its overall appeal. This also implies that the ad was strictly "informational" or fact-based.

To ensure that the evaluations did not simply reflect information in short-term working memory, a distractor task was introduced prior to administering the dependent measures in Stage 1. The participants were simply asked to "... list the brand names of cassette players you remember." Immediately following the distractor task, the subjects completed global evaluation scales. Also, they were asked to evaluate each of 15 attributes, and indicate their beliefs about whether the product actually had the attributes in question. Finally, the subjects were asked to report their purchase intentions.

Both experimental phases took a total of approximately 25 minutes to complete in Stage 1, but only 20 minutes in Stage 2. The shorter length of time in Stage 2 can be attributed to the fact that respondents did not again receive the attribute information about the DAT-111.

Measures

Mood. To assess the success of the manipulation, mood states were measured using the four-item Mood Short Form (MSF) scale (Peterson and Sauber 1983). The subjects were asked to indicate their feelings on the following five-point Likert-type scales (1-Strongly agree to 5-Strongly disagree): "At this moment I feel edgy or irritable," "For some reason, I am not very comfortable right now," "As I answer these questions, I feel very cheerful," and "Currently I am in a good mood." Ratings of the last two items were reverse-coded prior to data analysis. Consistent with the pretest results, subjects who read the stories designed to induce a positive mood were significantly happier for both Stage 1 and Stage 2 (x=4.02 and 4.03) than those who read the sad stories (x=2.68 and 2.49, p's <.001). The Cronbach's alpha coefficient of reliability for this four component measure of mood was .81 for Stage 1 and .84 for Stage 2. Although not initially considered, the intensity of the positive and negative mood states was approximately the same in both stages.

Global Evaluation. Respondents were asked to evaluate the DAT on five seven-point dependent measures (extremely low appeal-extremely high appeal, bad-good, unpleasant-pleasant, unagreeable-agreeable, unsatisfactory-satisfactory; a=.94) representing global evaluation of the DAT. The items used are similar to those used by Marks and Kamins (1988). The average of these five measures was used for data analysis.

Attitude Toward the Object. Attitude toward the object (Ao) was based on the summed set of beliefs (1-Strongly agree to 7-Strongly disagree) about the DAT's 15 attributes weighted by the evaluation (1-Very bad to 7-Very good) of these attributes.

Purchase Intentions. Purchase intentions were measured as an average of two five-point Likert-type scales (1-Definitely will buy to 5-Definitely will not buy; 1-Definitely like to have to 5-Definitely not like to have; Jamieson and Bass 1989). Ratings for the two purchase intentions scales were reverse-coded prior to data analysis.

RESULTS

Stage 1 results were analyzed using one-way multivariate analysis of variance (MANOVA). Three measures of interest (global evaluation, attitude toward the object, and purchase intentions) were analyzed simultaneously as a function of context-induced mood (positive, negative). Specifically, subjects in a positive mood provided different evaluations of the product on the four measures of interest than subjects in a negative mood (Wilks' L=.854; p<.007). To further clarify the nature of the results, supplementary analyses for each dependent variable were conducted separately. Not surprisingly, univariate tests confirmed the earlier MANOVA results. Relative to subjects in a negative mood those in a positive mood provided a more favorable global evaluation of the product (x= 3.25 vs. 3.89; F_{(1, 79)}=6.44, p<.01; eta²=.08) and a more favorable attitude toward the object (x= 608.34 vs. 506.10; F_{(1, 79)}=11.93, p<.001; eta²=.13). Finally, subjects in a positive mood provided higher purchase intentions for the product than those in a negative mood (x= 2.52 vs. 2.13; F_{(1, 79)}=5.75, p<.02; eta²=.07). Hypothesis 1, therefore, was supported.

TABLE

MEANS OF DEPENDENT MEASURES (STAGES 1 & 2)

Unlike previous research in the mood literature, this study also investigated the role of mood states in repeat evaluations. The Table shows the results on all dependent measures for each of the four groups (Positive-Positive, Positive-Negative, Negative-Negative, and Negative-Positive) for both stages. Repeated-measures MANOVA for each group revealed no significant differences in evaluations between Stages 1 and 2 (all p's=n.s.). Further analysis using paired sample t-tests also showed no significant differences for individual dependent measures (all p's=n.s.). The analysis performed here provides support for the hypothesized effects presented in Hypothesis 2.

DISCUSSION AND CONCLUSIONS

This study investigates the influence of context-induced mood states on initial and repeat product evaluations. Considered together, the results of the present study indicate that while mood states have a definite impact on initial product evaluations, repeat evaluations may not be susceptible to an individual's subjective mood state. This can appropriately be explained using the on-line and memory-based evaluation paradigms advanced in the literature.

According to past research, individuals can form evaluations of a stimulus in either of two processing paradigms. Consider first on-line processing. In this case, subjective affective states would influence judgments, since the judgments are being formed while individuals experience an affective subjective state (cf. Srull 1987). Our findings on the impact of affective subjective states on initial evaluations provide support for this. On the other hand, those individuals receiving the stimulus information without a specific objective to initially form an overall evaluation would not provide mood-laden judgments. The latter is true, since those judgments would be formed at a later time (perhaps, in the absence of mood). However, the impact of mood states on memory-based evaluations as described in the literature does not clearly account for our findings on repeat evaluations.

Repeat evaluations are those that follow already formed (or initial) evaluations of a stimulus. Unlike memory-based evaluations, repeat evaluations are, in fact, formed at an earlier stage (on-line) and occur again for a second time. Since individuals are not asked to process incoming attribute information when repeat evaluations take place, it is unlikely that affective states would have any impact on repeat evaluations. Indeed, the findings in this study provide support for this conceptualization. Unlike Srull's (1987) Experiment 2, in our study mood states were induced at both the encoding (initial) and retrieval (repeat) stages. Our findings are consistent, however, with Srull's results. This suggests that mood states may not be quite as salient in subsequent evaluations.

To put our findings in a proper perspective, some limitations of the study must be considered. First, the use of specific stories to induce the desired mood states does not imply that other contexts, or even other stories, can evoke similar moods. Therefore, evidence across studies using different contexts to manipulate mood may provide results that are either consistent or inconsistent with the findings of this study. Second, different kinds of positive and negative moods should be employed to test the prescribed hypotheses. Third, the study did not examine the information recalled by the respondents. According to earlier findings, the amount and type of information recalled by subjects were consistent with the mood congruency theory. It would be interesting, therefore, to examine whether recall of information in an experimental context (such as the one described in our study) would provide support for earlier findings in the literature. Finally, the study did not involve the manipulation of the affective tone of the ad. Past research has shown that the ad's affective tone may interact with context-induced mood to influence salient dependent measures (Gardner 1992). It would be useful to include such manipulations in future research.

The findings of the study have interesting implications for advertising, media planning, and retailing. Specifically, the results suggest that mood states are very important in forming initial evaluations of products. Therefore, a careful selection of the contexts within which ads and other promotional materials are presented is important. At the same time, the results suggest that attempts to influence mood states at later stages in an effort to change or enhance initial product evaluations (or product perceptions) may be unsuccessful. It would be beneficial, for example, to place major emphasis on the advertising of new products to influence initial, rather than later, judgments via mood states. This may be of particular importance to retail establishments that employ environmental stimuli (e.g., scents, background music) to influence mood states. Our results suggest that it may be costly and ineffective for retailers to influence evaluations of specific products.

REFERENCES

Batra, Rajeev and Douglas M. Stayman (1990), "The Role of Mood in Advertising Effectiveness," Journal of Consumer Research, 17(September), 203-214.

Bless, Herbert, Diane M. Mackie, and Norbert Schwarz (1992), "Mood Effects on Attitude Judgments: Independent Effects of Mood Before and After Message Elaboration," Journal of Presonality and Social Psychology, 63(4), 585-595.

Bower, Gordon (1981), "Mood and Memory," American Psychologist, 36(2), 129-148.

Bower, Gordon, Stephen Gilligan, and Kenneth Monteiro (1981), "Selectivity of Learning Caused by Affective States," Journal of Experimental Psychology: General, 110(December), 451-473.

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

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

Gardner, Meryl P. (1992), "Responses to Emotional and Informational Appeals: the Moderating Role of Context-Induced Mood States," in Advertising and Consumer Psychology, eds. E. Clark, T. Brock, and D. Stewart, Hillsdale, N.J.: Lawrence Erlbaum, Inc.

Goldberg, Marvin E. and Gerald J. Gorn (1987), "Happy and Sad TV Programs: How They Affect Reactions to Commercials," Journal of Consumer Research, 14(December), 387-403.

Hastie, Reid and Bernadette Park (1986), "The Relationship Between Memory and Judgment Depends on Whether the Judgment task is Memory-Based or On-Line," Psychological Review, 93(6), 258-268.

Isen, Alice (1970), "Success, Failure, Attention, and Reaction to Others: The Warm Glow of Success," Journal of Personality and Social Psychology, 15(4), 294-301.

Isen, Alice and Thomas Shalker (1982), "The Effect of Feeling State on Evaluation of Positive, Neutral, and negative Stimuli: When You 'Accentuate the Positive,' Do You 'Eliminate the Negative'?" Social Psychology Quarterly, 45(1), 58-63.

Jamieson, Linda F. and Frank M. Bass (1989), "Adjusting Stated Intention Measures to Predict Trial Purchase of New Products: A Comparison of Models and Methods," Journal of Marketing Research, 26(August), 336-345.

Johnson, Eric and Amos Tversky (1983), "Affect Generalization, and the Perception of Risk," Journal of Personality and Social Psychology, 45(1), 20-31.

Keller, Kevin L. (1991), "Cue Compatibility and Framing in Advertising," Journal of Marketing Research, 28(February), 42-57.

Knowles, Patricia A., Stephen J. Grove, and W. Jeffrey Burroughs (1993), "An Experimental Examination of Mood Effects on Retrieval and Evaluation of Advertisement and Brand Information," Journal of the Academy of Marketing Science, 22(2), 135-142.

Marks, Lawrence J., and Michael A. Kamins (1988), "The Use of Product Sampling and Advertising: Effects of Sequence of Exposure and Degree of Advertising Claim Exaggeration on Consumers' Belief Strength, Belief Confidence, and Attitudes," Journal of Marketing Research, 25(August), 266-281.

Mischel, W., Brian Coates, and Antonette Raskoff (1968), "Effects of Success and Failure on Self-Gratification," Journal of Personality and Social Psychology, 10(4), 381-390.

Mischel, W., Ebbesen, E., and A. Zeiss (1973), "Selective Attention to the Self: Situational and Dispositional Determinants," Journal of Personality and Social Psychology, 27, 129-142.

Moore, Bert, Bill Underwood, and D.L. Rosenhan (1973), "Affect and Altruism," Developmental Psychology, 8(1), 99-104.

Peterson, Robert A., and Matthew Sauber (1983), "A Mood Scale for Survey Research," in AMA Educators' Proceedings, eds. Patrick Murphy et al., Chicago, IL: American Marketing Association, 409-414.

Srull, Thomas K. (1987), "Memory, Mood, and Consumer Judgment," in Advances in Consumer Research, Vol. 14, eds. Melanie Wallendorf and P. Anderson, Provo, UT: Association for Consumer Research, 404-407.

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