An Extended Perspective on the Role of Emotion in Advertising Processing

Hyongoh Cho, University of Texas at Austin
Patricia A. Stout, University of Texas at Austin
ABSTRACT - The emotion construct in advertising processing continues to capture the attention of advertising researchers. This paper extends the perspective of the role of emotion in advertising processing in two ways: by integrating consumption characteristics or value-expressive beliefs into the role of emotional responses in the advertising context and by relating the emotional responses to a multidimensional construct of Aad. Three TV commercials representing a range of transformational and informational appeals are tested. Results indicate that certain emotional responses tend to exert greater influence on the affective dimension of Aad than the cognitive dimension of Aad. The relative impact of emotional responses on the value-expressive and utilitarian beliefs is contingent upon ad type. The impact of emotional responses on brand attitude is completely mediated by the multidimensional measures of Aad and Cb. Overall, the relative strength of the causal linkage between emotional responses and the multidimensional constructs of the mediating parameters depends on specific emotional characteristics as well as the type of ad execution.
[ to cite ]:
Hyongoh Cho and Patricia A. Stout (1993) ,"An Extended Perspective on the Role of Emotion in Advertising Processing", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 692-697.

Advances in Consumer Research Volume 20, 1993      Pages 692-697

AN EXTENDED PERSPECTIVE ON THE ROLE OF EMOTION IN ADVERTISING PROCESSING

Hyongoh Cho, University of Texas at Austin

Patricia A. Stout, University of Texas at Austin

ABSTRACT -

The emotion construct in advertising processing continues to capture the attention of advertising researchers. This paper extends the perspective of the role of emotion in advertising processing in two ways: by integrating consumption characteristics or value-expressive beliefs into the role of emotional responses in the advertising context and by relating the emotional responses to a multidimensional construct of Aad. Three TV commercials representing a range of transformational and informational appeals are tested. Results indicate that certain emotional responses tend to exert greater influence on the affective dimension of Aad than the cognitive dimension of Aad. The relative impact of emotional responses on the value-expressive and utilitarian beliefs is contingent upon ad type. The impact of emotional responses on brand attitude is completely mediated by the multidimensional measures of Aad and Cb. Overall, the relative strength of the causal linkage between emotional responses and the multidimensional constructs of the mediating parameters depends on specific emotional characteristics as well as the type of ad execution.

Since the early 80s, many advertising researchers have extended their attention from the cognitive structures of advertising (e.g., product beliefs, comprehension of ad message, source-related cognitive responses) to the role of the affective properties in advertising processing (e.g., emotional responses, Aad, mood, music). It is suggested that affect and cognition are separate and partially independent systems and that although they ordinarily function conjointly, affect may not require the mediation of cognitive processes as a necessary element in the generation of emotional states (Zajonc 1980). Attitude is not only mediated by salient beliefs about the product attributes, but also by mere association with the affective stimuli in the ad (Mitchell and Olson 1981).

Although the unique role of emotional responses in the consumer behavior has been widely investigated, it is our feeling that current knowledge about the emotion construct is still narrowly focused and discloses a wide gap between the emotional processing in the advertising context and the emotional characteristics in the consumption experience. This paper intends to integrate the consumption characteristics into the role of emotional responses in the advertising context. More specifically, we examine the impact of the emotional responses on attitude toward the ad (Aad), brand cognition (Cb), and brand attitude (Ab) and attempt to relate various consumption experiences and motives to these relationships.

THE ROLE OF EMOTIONAL RESPONSE IN ADVERTISING PROCESSING

The relationship between emotional response and Aad

Extensive research has investigated the causal linkage between a variety of emotional properties and Aad. For instance, Aad can be influenced by various emotional properties such as PAD (or Pleasure, Arousal, and Dominance) dimensions (Holbrook and Batra 1987), discriminant negative and positive feelings (Edell and Burke 1987), various depth of emotional associations with oneself (Stout and Leckenby 1986), emotional intensity (Stephens and Russo 1987) and warmth (Stayman and Aaker 1988). Despite some conceptual ambiguity between emotional responses (ERs) and other types of affective constructs, it appears that ERs are not only distinct from Aad and other types of affect but play a discriminant role in the formation of Aad.

However, previous research tended to oversimplify the Aad construct by conceptualizing it as a unidimensional construct and measuring it as the overall evaluation of the ad stimuli. Many studies have evidenced the multidimensional Aad construct consisting of cognitive and affective components (Burton and Lichtenstein 1988; Madden et al. 1988; Miniard et al. 1990; Olney et al. 1991). The multidimensional construct is not only superior to the unidimensional approach in predicting Ab, but provides in-depth information on the dynamic relationship between the Aad and other competing parameters.

While few studies investigated how ERs relate to the multidimensional construct of Aad, the incorporation of ERs into the multidimensional Aad construct is expected to enrich our understanding about the characteristics of the affective properties in ad processing. A dual processing perspective for the formation of Aad may provide an useful framework underlying this relationship. Several studies propose that Aad can be formed via central and peripheral processing (MacKenzie and Lutz 1989; Miniard et al. 1990). Although ERs have not been directly incorporated in their models, ERs have been characterized as an affective construct and are likely to represent the affective dimension of the Aad. Based on the multidimensional perspective of the formation of Aad and dual processing perspectives, we hypothesize that:

H1: ERs exert significant influence on the formation of Aad. The impact of ERs on the affective dimension of Aad is greater than their impact on the cognitive dimension of Aad.

The relationship between emotional response and brand cognition

While many studies postulated ERs as an antecedent of Aad and Ab, few studies have studied the causal linkage between ERs and Cb (Burke and Edell 1989; Edell and Burke 1987). Even those studies which found a significant relationship between ERs and Cb did not provide an adequate explanation about the underlying mechanism for the transfer of ERs to Cb. The insufficient theoretical support for this relationship in the stream of Aad research may be partly attributed to the widespread perception that ERs are likely to represent and facilitate heuristic (Chaiken 1980) or peripheral processing (Petty and Cacioppo 1981). ERs can occur even in the absence of minimal cognitive processing and tend to be automatic or involuntary (Zajonc 1980).

Contrary to the previous emotion studies, we suggest that ERs can influence the evaluation of the product attributes for three reasons. First, the affective and cognitive elements of the persuasion process may be highly intertwined and have a synergistic effect on consumers' information processing. ERs not only affect the amount of cognitive processing, but also increase the accessibility of certain types of product attributes congruent with their emotional state (Forgas and Bower 1987). ERs also have a direct and an indirect impact on brand beliefs via ad judgment (Burke and Edell 1989; Edell and Burke 1988). Thus, it appears that various types of ERs can enhance or decrease the evaluation of product information, creating discriminant impacts on Cb. Secondly, previous emotion and Aad studies tended to narrowly define the boundary of Cb. Most studies have focused on only functional, tangible or utilitarian characteristics of product attributes in the measure of Cb or the development of test ads. Considering that many researchers have distinguished between utilitarian and value-expressive functions of a product (Hirschman and Holbrook 1982; Howard and Sheth 1969), it is rather surprising that the value-expressive characteristics of product attributes have been largely ignored in advertising processing. The incorporation of image benefits or value-expressive attributes to the multiattribute measure of brand beliefs demonstrated significant incremental contribution of Cb to Ab (Mittal 1990). Finally, the extension of the consumer research from buying decision to consumption experience has enriched our understanding about the importance of value-expressiveness or imagery in the consumption process. The value-expressiveness of product consumption has been reflected in the studies of hedonic consumption (Hirschman and Holbrook 1982), meaning of the possession (Prentice 1987), consumption motives (McGuire 1976), symbolism in the product (Belk 1982), and self-concept (Sirgy 1982).

The value-expressive and utilitarian attributes can be characterized by the tangibility of the product attributes, type of benefits, and drives underlying the consumption. Value-expressive attributes are directed at abstract, subjective or intangible aspects of a product whereas the utilitarian attributes derive from the concrete, objective or tangible characteristics of a product (Fin 1985). Value-expressive attributes are targeted at the aesthetic or symbolic benefits as a means to express valued personal traits, whereas utilitarian attributes contribute to the instrumental or functional benefits as a means to affect and control the physical environment (Prentice 1987; Swan and Combs 1976). Finally, the product attributes can be characterized by either hedonically driven consumption motives or economic, utilitarian motives to manage the surrounding environment (Hirschman and Holbrook 1982; McGuire 1976).

Overall, the emotional components in the consumption experience are inseparable from self-expressiveness of oneself, and when consumers look for emotional experience in the product, they are likely to become more receptive to the value expressive attributes in the product as opposed to utilitarian attributes of the product. Therefore, it is our contention that the ERs created by ad exposure will stimulate and facilitate the processing of value-expressive and imagery characteristics of product attributes. We hypothesize that :

H2: ERs are a significant predictor of brand cognition. The effect of the ERs on the value-expressive responses is greater than their effect on the utilitarian responses.

The direct impact of ERs on Ab beyond the mediation of Aad and Cb

Previous studies demonstrate somewhat conflicting findings regarding the direct impact of ERs on Ab beyond the mediation of Aad. Many studies find that the impact of ERs on Ab becomes insignificant when Aad is incorporated in the equation, suggesting the path between ERs and Ab as being completely mediated by Aad (Batra and Ray 1986; Edell and Burke 1987; Holbrook and Batra 1987). In contrast, Stayman and Aaker (1988) found that some of the feeling responses were not completely mediated by Aad, indicating the direct impact of the feeling responses on Ab. They explain that "the single exposure typically used in previous studies may have been insufficient to obtain either the necessary variation in feeling response or the the association between the response and attitudes" (p. 371). More conclusive evidence for the direct impact of ERs on Ab beyond Cb and Aad is provided by Burke and Edell (1989). While each dimension of feeling responses had direct and indirect impact on Aad and Cb via ad judgment response, upbeat and negative feelings had a direct impact on Ab beyond the mediation by either Aad or Cb. Therefore, it is likely that a wide range of ERs is a necessary condition for its direct impact on Ab beyond Aad.

Considering that these findings are based on unidimensional constructs of Aad and Cb, it is unclear whether the direct effect of ERs on Ab would be still significant after multidimensional measures of Aad and Cb are introduced. Since the global evaluation of Aad would explain less variance of Ab than the multidimensional measures of Aad, it is likely that the direct impact of ERs on Ab decreases when Aad is operationalized as a multidimensional construct. Also, some variance of Ab directly explained by ERs may be captured by the multidimensional measures of Cb. We hypothesize the following:

H3: The direct impact of the ERs on Ab with unidimensional measures of Aad and Cb becomes insignificant when both Aad and Cb are measured as a multidimensional construct.

METHOD

Subjects: Fifty five student subjects were recruited from undergraduate communication courses at a large southwestern university and given course credit for participation.

Stimuli: The test ads include three 30-second TV commercials that have not been aired previously in the area and are for brands not available in the area to avoid the confounding effect of prior Ab with the ad-generated responses. The three products include beer, toilet tissue, and insect repellent, typically used by the target subjects.

Since we intend to test the causal linkage between ERs and other types of cognitive and affective parameters (i.e., value-expressive vs. utilitarian brand beliefs, affective vs. cognitive dimensions of Aad), it is important to extend the opportunity to generate a wide range of cognitive and affective responses to ad. We employed the informational/transformation scales of Puto and Wells (1984) to select the ads representing both emotional and informational appeals. Based on a pretest of eighteen ads, three 30-second commercials were selected. The test ads include 1 primarily transformational ad (t value= 3.63, p <.01), 1 primarily informational ad (t value= -8.75, p <.001), and 1 equally transformational/informational ad (t value= .20, p= .844).

Procedures: Subjects were instructed to view each of the three videotaped commercials. The sequence of the commercials was equally randomized to minimize any order effect. After the subjects viewed each videotaped commercial three times, they were asked to indicate their ERs, Aad, Cb, and Ab for each of the test commercials.

Measures

ERs: The emotion checklist consisting of 50 adjective items was developed on the basis of the analysis of the previous emotion measures (Edell and Burke 1987; Holbrook and Batra 1987; Westbrook and Oliver 1991). Holbrook and Batra's (1987) 7-point check-mark scales are employed to measure the emotion scales with instructions on "how the experience of watching the ad affected you emotionally, not how you describe the ad."

Aad: Sixteen 7-point semantic items were extracted from the previous Aad studies which successfully captured both affective and cognitive components of Aad in the measures (Burton and Lichtenstein 1988; Madden et al. 1988; Olney et al. 1990).

Cb: Subjects were asked to write down any thoughts, feelings or images toward the brand that went through their mind while watching the commercial. In addition, they were instructed to indicate the intensity of the valence for each thought by putting a number ranging from +3 to -3 with +3 representing a very positive thought and -3 a very negative thought. It is believed that the intensity of the valence may vary across different thoughts and by allowing unequal weight for each thought we may be able to measure the subjects' Cb more accurately. Two expert judges were employed to encode the subjects' free thoughts into the utilitarian and value-expressive responses toward the brand and commercial-related thoughts. Brand-related responses which may not be interpretable as either utilitarian or value-expressive responses are treated as general responses to the brand.

Ab: Ab was operationalized by three 7-point semantic differential scales. The scales of Ab include like/dislike, favorable/unfavorable, good/bad.

DATA ANALYSIS AND DISCUSSION

First, the emotion scales and the Aad scales were factor analyzed. In the emotion scales, 7 principle components with an eigenvalue greater than 1 emerged, far too many to use in the present analysis. A further examination of the scree test indicates that three factors are clearly distinguished from the rest. Accordingly, we extracted three principal components, all with eigenvalues greater than 4, accounting for 61.5 percent of the variance in the emotion scores. Due to the page limitation, loadings on the resulting dimensions using varimax-rotation are not reported here. Forty of 50, or 80 percent, of the emotion scales have correlated loadings greater than 0.50 on the three factors.

The structure of the emotional dimensions strongly supports the feeling scales identified in the studies of Edell and Burke (1987; Burke and Edell 1989). We refer to these three dimensions of feeling scales as upbeat, warm, and irritated feeling responses. In fact, these dimensions are not totally in conflict of the traditional PAD model. The Arousal dimension is well reflected in the upbeat scales, whereas the Pleasure dimension may not be a unidimensional construct, but rather consists of two distinct dimensions; that is, warm feelings and irritated feelings. Previous studies show that the positive and negative emotional responses can occur simultaneously and do not necessarily share a unidimensional structure (Burke and Edell 1989; Edell and Burke 1987). The absence of the dominance dimension may be attributed to the limited set of the test commercials. Alternatively, it may be argued that dominance dimension is not an important aspect of the consumers' ERs to TV commercials in general. The coefficient alpha for each scale is .97 for warm, .95 for upbeat, and .92 for irritated feelings.

The principal component analysis for the Aad measures strongly supports two distinct dimensions with eigenvalue greater than 1, accounting for 68.2 percent of the variance. Fourteen of 16 items have high loadings greater than .5 on these two dimensions. The structure of these two dimensions are virtually identical to the affective and cognitive dimensions developed in previous research. While the third dimension may be interpreted as "interestingness", as suggested by Olney et al. (1990), it is not included in our study due to the small amount of the variance explained by this dimension and the inadequate theoretical explanation for the dimension. The scalar items highly loaded on the affective dimension include affectionate/not affectionate, coldhearted/warmhearted, depressing/uplifting, fun to watch/not fun to watch, pleasant/unpleasant, soothing/not soothing, tasteful/tasteless and unattractive/attractive. The adjectives indicating the cognitive dimension are believable/unbelievable, convincing/not convincing, helpful/not helpful, ineffective/effective, informative/uninformative and persuasive/nonpersuasive. The coefficient alpha for each scale is .95 for the affective dimension, and .90 for the cognitive dimension. Three emotion scales and two Aad scales were formed by summing the ratings for those items that loaded greater than 0.5 on a factor, reverse scoring where necessary.

After coding the open-ended question for Cb, the interjudge reliability for the value-expressive and utilitarian measures of Cb was 83 percent. Overall, 86 percent of brand-related responses were either "utilitarian" or "value-expressive" responses, leaving the remaining 14 percent under a "general response" category. For the subsequent analysis, the scores of the utilitarian and value-expressive responses are computed by counting the number of value-expressive and utilitarian thoughts and multiplying them by the intensity of the valence for each thought. In the other dependent measures, the coefficient alphas for global Aad and Ab are .96 and .96 respectively.

A separate analysis for each commercial is conducted to examine the relationship between the ERs and the subsequent parameters because averaging the scores across three commercials may confound the idiosyncratic effect of the ad executions with the impact of emotional responses on the other explanatory variables. LISREL modeling (Joreskog and Sorbom 1986) is employed to test the causal linkage among ERs, Aad, Cb and Ab. In the model testing, ERs are postulated as having direct and indirect impact on Ab via multidimensional Aad and two components of Cb. Each of the multidimensional components of Aad and Cb is also hypothesized to have a direct effect on Ab (see Figure 1).

To conserve space, no further discussion on the latent measurement statistics is reported. In a sense, the proposed model does not exclude any alternative models; rather it mainly tests the relationship between ERs and the other explanatory variables of the interest. The proposed model fits the data fairly well for all three commercials in terms of chi-square and Joreskog and Sorbom's (1986) goodness-of-fit index. While chi-square may not be a reliable measure for a small sample size, the satisfactory GFI indicates a relatively good fit in the proposed model. Table 1 contains the path coefficients resulting from Generalized Least Square estimation of the model parameters.

Relationship between ERs on Multidimensional Aad

It is hypothesized that the impact of ERs on the affective dimension of Aad is greater than their impact on the cognitive dimension of Aad. The standardized path coefficients of the equation show that warm feelings have a significant effect only on the affective dimension of Aad across all three commercials, whereas upbeat and irritated feelings tend to exert significant influence on both affective and cognitive dimensions of Aad depending on the type of ad execution. More specifically, upbeat feelings demonstrate a strong impact on the affective Aad in the trans/informational ad, whereas the relationship becomes significant only through the cognitive dimension of Aad in the transformational ad. No impact of upbeat feelings on the multidimensional Aad is found for the informational ad type. On the other hand, irritated feelings become a salient indicator of the affective Aad in the trans/information and transformational ads. For the informational ad, irritated feelings influence both affective and cognitive components of Aad. Although the mechanism underlying the transfer of the upbeat and irritated feelings to the cognitive dimension of Aad is not known, our findings support the study of Burke and Edell (1989) in that upbeat and negative feelings demonstrate strong influence on the evaluation judgment which can be interpreted as the cognitive dimension of the Aad.

Overall, H1 is strongly supported for warm feelings and partially supported for irritated feelings. However, the relationship between ERs and multidimensional Aad appears to be far more complex than hypothesized, indicating that the relative strength of the causal linkage between the ERs and the multidimensional measures of Aad may be contingent upon specific emotional characteristics as well as the executional types.

FIGURE 1

PROPOSED MODEL ON THE ROLE OF EMOTION IN ADVERTISING PROCESSING

Effects of ERs on Multicomponents of Cb

As predicted in H2, ERs are found to be an important indicator of multidimensional Cb. As a further step, the impact of ERs on Cb is hypothesized to become salient via the effect on the value-expressive Cb. This hypothesis is partially supported for the trans/informational ad in that only value-expressive component of Cb is significantly influenced by upbeat and irritated feelings. For the informational ad, upbeat feelings have a substantial impact on value-expressive Cb (p=.10) whereas irritated feelings become a significant determinant of both value-expressive and utilitarian Cb. On the contrary, no direct relationship between ERs and multidimensional Cb for either individual or jointed ERs is observed in the transformational ad. It appears that ERs generated by a transformational appeal tend to activate the salience of the executional features in the ad without influencing brand-related responses. Similarly, warm feelings do not have any significant impact on either value-expressive or utilitarian Cb across various types of ad execution, indicating that warm feelings represent solely execution-based peripheral processing.

In sum, the effect of warm feelings on Ab is completely mediated by the Aad construct. On the other hand, upbeat feelings are likely to interfere with the value-expressive attribute processing, whereas irritated feelings tend to exert influence on both value-expressive and utilitarian Cb depending on ad type. It may be argued that the same type of emotional responses leads to various communication outcomes depending on the executional elements which characterize the emotional reaction. The complexity of the relationship between ERs and multidimensional Cb across various types of the ad exceeds the initial expectation and requires more discrete analysis in the future. Nevertheless, considering that little effort has been given to the relationship between ERs and the multidimensional construct of Cb, this preliminary conceptualization may be considered as an initial step toward a comprehensive framework for the ad processing involving multidimensional constructs of the mediating parameters.

Direct Effects of ERs on Ab

H3 predicts a significant decrease in the direct impact of the ERs on Ab when both Aad and Cb are measured as a multidimensional construct. None of the individual ERs or their joint effects are significant across the three commercials. The result indicates that the impact of the ERs on Ab is completely mediated by the multidimensional measures of Aad and Cb. In order to examine whether it holds true for the unidimensional constructs of Aad and Cb, a separate analysis with a unidimensional design was conducted. The result from the reduced unidimensional model shows that upbeat feelings have a direct impact on Ab (path coefficient= .48, p<.05) in the transformational ad whereas the direct path ERs-Ab is not significant for either the individual ERs or their joint effects in the other types of the ads. The direct effect of upbeat feelings on Ab beyond the global measures of Aad and Cb is consistent with previous findings (Burke and Edell 1989). Thus, this result supports H3 for the transformational ad in that certain ERs which directly influence Ab become insignificant after multidimensional measures of Aad and Cb are introduced.

A further look at the multidimensional constructs shows that the effect of warm feelings on Ab is mainly mediated by affective dimension of Aad regardless of various types of ad execution. Upbeat feelings tend to influence Ab through its impact on the multicomponents of Aad and value-expressive Cb. On the other hand, the impact of irritated feelings on Ab appears to be mediated by both affective/cognitive Aad and value-expressive/utilitarian components of Cb depending on the type of ad execution. Therefore, warm feelings appear to activate mainly a peripheral processing route by influencing the affective dimension of Aad, whereas upbeat and negative feelings are likely to facilitate both central and peripheral processing by affecting both multidimensional components of Aad and Cb.

TABLE 1

PARAMETER ESTIMATES FOR THE CAUSAL LINKAGE

This study attempted to extend the role of emotion in the ad processing by incorporating ERs and multidimensional constructs of Aad and Cb. In an attempt to elaborate on the causality among these extended constructs, three specific hypotheses were presented. Overall, the extended model reveals rich diagnostic information about how ERs interact with the mediating variables toward the formation of brand attitude. Yet this study is not without limitations. Due to the small sample size, the explanatory power of this study is certainly limited. Although we attempted to include different types of appeals, the limited set of TV commercials may not have captured an adequate range of emotional responses, limiting our ability to test certain relationships in the model. While different ad processing mechanisms may exist across various types of ad appeals, we failed to find any consistent pattern in this regard. A more discrete approach is likely to enrich our understanding about the role of ERs in ad processing since the relative impact of ERs on the subsequent parameters may be contingent upon specific emotional factors as well as ad characteristics.

REFERENCES

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

Belk, Russell (1982), "Developmental Recognition of Consumption Symbolism," Journal of Consumer Research, Vol.9 (June), 4-17.

Burke, Marian C., and Julie A. Edell (1989), "The Impact of Feelings on Ad-Based Affect and Cognition," Journal of Marketing Research, 26 (February), 69-83.

Burton, Scot and Donald R. Lichtenstein (1988), "The Effect of Ad Claims and Ad Context on Attitude Toward the Advertisement," Journal of Advertising, 17 (1), 3-11.

Chaiken, Shelly (1980), "Heuristic Versus Systematic Information Processing and the Use of Source Versus Message Cues in Persuasion," Journal of Personality and Social Psychology, Vol. 39, No. 5, 752-766.

Edell, Julie A. and Marian C. Burke (1987), "The Power of Feelings in Understanding Advertising Effects," Journal of Consumer Research, 14 ( December), 421-433.

Fin, Adam (1985), "A Theory of the Consumer Evaluation Process For New Product Concepts," in Jagdish N. Sheth, eds., Research in Consumer Behavior, Vol. 1, Greenwich, CT: JAI Press, 35-65.

Forgas, Joseph P. and Gordon H. Bower (1987), "Mood Effects on Person-Perception Judgments," Journal of Personality and Social Psychology, Vol. 53, No. 1, 53-60.

Hirschman, Elizabeth and Morris B. Holbrook (1982), "Hedonic Consumption: Emerging Concepts, Methods and Propositions," Journal of Marketing, 46 (Summer), 92-101.

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

Howard, John A. and Jagdish N. Sheth (1969), The Theory of Buyer Behavior. New York: Wiley.

Joreskog, Karl G. and Dag Sorbom (1986), LISREL VI: Analysis of Linear Structural Relationships by the Method of Maximum Likelihood. Mooresville, IN: Scientific Software, Inc.

MacKenzie, Scott B. and Richard J. Lutz (1989), "An Empirical Examination of the Structural Antecedents of Attitude Toward the Ad in an Advertising Pretesting Context," Journal of Marketing, 53 (April), 48-65.

Madden, Thomas J., Chris T. Allen and Jacquelyn L. Twible (1988), "Attitude Toward the Ad : An Assessment of Diverse Measurement Indices Under Different Processing "Sets"," Journal of Marketing Research, 25 (August), 242-252.

McGuire, William J. (1976), "Some Internal Psychological Factors Influencing Brand Choice," Journal of Consumer Research, 2 (March), 302-319.

Miniard, Paul W., Sunil Bhatla and Randall L. Rose (1990), "On the Formation and Relationship of Ad and Brand Attitudes : An Experimental and Causal Analysis," Journal of Marketing Research, 27 (August), 290-303.

Mitchell, Andrew A. and Jerry C. Olson (1981), "Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude?," Journal of Marketing Research, 18, 318-332.

Mittal, Banwari (1990), "The Relative Roles of Brand Beliefs and Attitude Toward the Ad as Mediators of Brand Attitude : A Second Look," Journal of Marketing Research, 27, 209-219.

Olney, Thomas J., Morris B. Holbrook and Rajeev Batra (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.

Petty, Richard E. and John T. Cacioppo (1981), Attitudes and Persuasions: Classic and Contemporary Approaches, Dubuque, IA: William C. Brown.

Prentice, D. (1987), "Psychological Correspondence of Possessions, Attitudes and Values," Journal of Personality and Social Psychology, 53 (6), 993-1003.

Puto, Christopher P. and William D. Wells (1984), "Informational and Transformational Advertising: The Differential Effects of Time," in Advances in Consumer Research, 11, ed. Thomas C. Kinnear, Provo, UT: Association for Consumer Research, 572-576.

Sirgy, M. Joseph (1982), "Self-Concept in Consumer Behavior: A Critical Review," Journal of Consumer Research, 9, 287-300.

Stayman, Douglas M. and David A. Aaker (1988), "Are All the Effects of Feelings Mediated by Aad?," Journal of Consumer Research, 15 (December), 368-373.

Stephens, Debra L. and J. Edward Russo (1987), "Predicting Post-Advertising Attitudes," working paper, University of Maryland.

Stout, Patricia A., and John D. Leckenby (1986), "Measuring Emotional Response to Advertising," Journal of Advertising, 15, (4), 35-42.

Swan, John E. and Linda J. Combs (1976), "Product Performance and Consumer Satisfaction: A New Concept," Journal of Marketing, 40, (April), 25-33.

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

Zajonc, Robert B. (1980), "Feeling and Thinking: Preferences Need No Inferences," American Psychologist, 35 (February), 151-175.

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