Role of Attitude Toward Brand Advertising on Consumer Perception of a Brand Extension

ABSTRACT - This paper reports findings from an experiment that investigates the role of consumers' attitude toward the brand advertising (Aad) on the accessibility and perceived appropriateness of the brand attributes in their perception of a brand extension. The effect is examined either for a comparable extension (sneakers from denim clothing) or for a moderately noncomparable extension (perfume/cologne from denim clothing). Findings suggest that positive Aad significantly increases both the accessiblity and the perceived appropriateness of the brand attributes. The effect on the accessibility of the brand attributes is observed in a similar pattern for both comparable and moderately noncomparable extensions. The effect on the perceived appropriateness of the brand attributes occurs more consistently for a moderately noncomparable extension than for a comparable extension. Additional analysis indicates that Aad strongly influences consumers' overall attitude toward a brand extension, with the effect being greater for a moderately noncomparable extension than for a comparable extension.

Citation:

Jung S. Lee (1995) ,"Role of Attitude Toward Brand Advertising on Consumer Perception of a Brand Extension", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: .

Advances in Consumer Research Volume 22, 1995      Pages 116-122

ROLE OF ATTITUDE TOWARD BRAND ADVERTISING ON CONSUMER PERCEPTION OF A BRAND EXTENSION

Jung S. Lee, Towson State University

ABSTRACT -

This paper reports findings from an experiment that investigates the role of consumers' attitude toward the brand advertising (Aad) on the accessibility and perceived appropriateness of the brand attributes in their perception of a brand extension. The effect is examined either for a comparable extension (sneakers from denim clothing) or for a moderately noncomparable extension (perfume/cologne from denim clothing). Findings suggest that positive Aad significantly increases both the accessiblity and the perceived appropriateness of the brand attributes. The effect on the accessibility of the brand attributes is observed in a similar pattern for both comparable and moderately noncomparable extensions. The effect on the perceived appropriateness of the brand attributes occurs more consistently for a moderately noncomparable extension than for a comparable extension. Additional analysis indicates that Aad strongly influences consumers' overall attitude toward a brand extension, with the effect being greater for a moderately noncomparable extension than for a comparable extension.

INTRODUCTION

The purpose of this paper is to investigate the role of consumers' attitude toward the brand advertising on the way they relate a brand extension to the brand. Although the role of brand concept or brand attitude on consumer perception of a brand extension has been actively pursued during the past several years, few studies have investigated the impact of brand advertising on the process. Brand attitude and brand concept are relatively stable, yet they are affected by the current marketing communication environment. Consumers' response to brand advertising is likely to have a significant influence on their attitude toward the brand and the brand-salient concepts at the time of a brand extension, which, in turn, would influence their perception of the brand extension. Furthermore, it may have a direct effect, in addition to its mediated effect through the brand, to their perception of a brand extension.

It is reasonable to think that marketers try a brand extension instead of introducing a new brand, if they believe that consumers will relate a brand extension to their existing positive attitude and image of the brand. For example, the marketer of Ivory shampoo would hope that consumers who like Ivory soap would also like Ivory shampoo and associate it with the 'pureness' of Ivory soap. Thus, in general, the more consumers can associate a brand extension to the existing brand, the better it is.

Among different ways consumers relate a brand extension to the brand, the amount of brand attributes that become accessible when they think of a brand extension, and the degree to which those attributes are perceived to be appropriate for a brand extension are of special interest in this study. This is because they seem to represent two distinctively different ways consumers relate a brand extension to the existing brand.

Several studies in brand extension suggest that consumers' attitude toward the brand influences their overall evaluation of a brand extension through a process of 'affect transfer' when a brand extension is made to a highly similar category (e.g., Aaker and Keller, 1990; Park et al., 1991). However, few studies have investigated whether the overall attitude toward the brand (or toward brand advertising) influences the way consumers relate specific brand attributes to a brand extension.

Research on the role of affect on cognitive organization suggest that people's affective state (either from the context or from the stimuli) significantly influences the way they organize incoming information (e.g., Isen and Daubman, 1984; Murray et al., 1990). Research on affective response to advertising has shown that (1) attitude toward the ad indicates in part consumers' affective response to the ad (Batra and Ray, 1986), and that (2) a brand name, when cued later, can retrieve the affect associated with its advertising (Edell and Burke, 1992; Keller, 1987; Stayman and Batra, 1991). From this line of reasoning, it is speculated that the consumers' attitude toward the brand advertising may influence the way they make a sense of a brand extension later. Although the conceptual distinction between affect as a general feeling state and attitude as an overall evaluative predisposition is important (Cohen and Areni, 1990; Isen, 1984), it would be interesting to see if attitude toward the brand advertising (as an indicator of an overall affective response) has a significant impact on consumers' cognitive organization of brand attributes in a brand extension. It is not the purpose of the study to determine whether the effect is a consequence of pure affect or of attitude. Thus, the term 'affect associated with the ad' and 'attitude toward the ad' are used interchangeably in the current study.

Affect and Cognitive Organization

Research have shown that positive affect influences people's cognitive organization (see Forgas, 1991, for a review). However, there is little consensus about what these effects are and how they occur. First, the state-dependency model states that people in a positive state are more likely to selectively activate, attend to, store, and retrieve positive materials than those in a neutral mood are (Forgas and Bower, 1987). Second, the cognitive flexibility model states that positive affect facilitates people's access to both "proattitudinal" and "counterattitudinal" messages, depending on the goal of processing (Murray et al., 1990). This is because positive affect enhances cognitive flexibility, instead of state dependency. The Cue-accessibility model proposes that positive affect cues more material and more diverse material compared to neutral affect (Isen and Daubman, 1984).

Less is known about the role of negative affect on cognitive organization. Studies have shown that people under a negative mood engaged in more critical and systematic information processing, had a narrower latitude of acceptance, and formed a less inclusive category (see Schwartz and Bless, 1991 for a review). On the other hand, it has been also suggested that individuals' information processing might become less critical or systematic under a negative feeling state than under a neutral state (Cialdini, Darby and Vincent, 1973; Isen and Daubman, 1984). This reportedly is because people engage in an affect repair process when they experience a negative feeling.

Affect and biased inference-making

Other research has shown that positive affect influences the nature of inferences people make about given information, rather than its organization. For example, people in a positive state rated the same traits of a target person in a more positive way than those in a neutral state did (Kunda, 1987). It was suggested that this biased inference was due to "misattribution" of positive affect from the mood-eliciting stimuli to the object being evaluated (Schwarz and Clore, 1983). If misattribution is the underlying mechanism, it is likely to occur consistently under both positive and negative affect.

Moderated by Task Complexity

A question may be raised as to whether positive affect always influences cognitive organization or inference-making. Schwarz and Clore (1988) noted that judgments that were based on often-rehearsed and highly familiar information ("crystalized") were much less likely to be influenced by transient mood than were new, ambiguous, or difficult judgments. This reportedly is because there is less need for a complex cognitive context when the task is simple (Branscombe and Cohen, 1991). It has also been argued that the relationship between affect and cognitive organization is often meaning-specific. That is, affect influences the organization of the materials that are meaningfully related to the type of affect (see Isen, 1987, for a review). This finding may be applied to the role of affect on biased inference-making. That is, affect is most likely to bias the inference-making for those concepts that are meaningfully related to the object being evaluated.

When an evaluation task itself is complex, however, it becomes difficult to judge the relevance of the accessible concepts to the object being evaluated. Thus, affect may bias the inference-making rather consistently for all concepts activated in the context. For a relatively easy task, on the other hand, the effect is likely to occur to a greater degree for those concepts that are relevant to the object being evaluated. Thus, task complexity may moderate the influence of affect on biased inference-making, because it influences judgment on the concept relevance, which, in turn, moderates the extent of biased inference-making.

Implications for consumer perception of a brand extension

The findings on the role of affect on cognitive organization and biased inference-making seem to provide useful insights on the investigation of the relationship between consumers' attitude toward the brand advertising and their perception of a brand extension. First, findings on affect and cue-accessibility imply that consumers' attitude toward the brand advertising may influence the accessibility of the brand attributes when they think of a brand extension. Second, the discussion on biased inference-making suggests that attitude toward the brand advertising may also affect the perceived appropriateness of the brand attributes in a brand extension, as a specific type of inferences.

Furthermore, the extent of the effect may vary across different types of brand extensions. Past research in brand extension has defined an extension product as being either similar or dissimilar to the core product. The current study adopts a different view in which many extensions could be simply noncomparable to the brand to the ordinary consumers, unless external cues are provided that clarify the relationship. Thus, there could be comparable brand extensions that are easy to judge their relationship with the brand, and noncomparable brand extensions of which the judgments are more difficult.

Johnson (1984) defines product comparability as the degree to which two products are represented by the same set of attributes. Thus, a comparable extension is one that is evaluated by the same set of attributes as those of the core brand, although the relative salience of each attribute may vary between the two. A moderately noncomparable extension, on the other hand, refers to one that shares only a few of its salient attributes with those of the brand, while each product possesses unique attributes that are irrelevant to the other. A completely noncomparable extension is excluded in this study because it is unlikely for this type of extension to occur in a real situation.

HYPOTHESES

Based on the above discussion, the following hypotheses are made about the effect of consumers' attitude toward the brand advertising (Aad) on the accessibility and perceived appropriateness of the brand attributes in their perception of a brand extension.

On Brand-Attribute Accessibility

H1a: The more positive consumers' attitude toward the brand advertising (Aad) is, the higher the accessibility of brand attributes is in their perceptions of a brand extension.

H1b: The effect of Aad on the brand-attribute accessibility is stronger for a moderately noncomparable extension than for a comparable extension.

On Perceived Appropriateness of Brand Attributes

H2a: The more positive consumers' attitude toward the brand advertising (Aad) is, the more they are likely to think that the brand attributes are appropriate for a brand extension.

H2b: The effect of Aad on the perceived appropriateness of the brand attributes is more consistent for a moderately noncomparable extension than for a comparable extension.

METHODS

Overview

Fictitious print advertisements were prepared for denim clothing (women's jeans, men's jeans, jackets). Two-thirds of the participants saw the ads and the remaining one-third were given the brand names only to form a baseline measure. Each participant was then asked questions about a future brand extension that is either comparable (sneakers) or moderately noncomparable (perfume/cologne) to denim clothing.

Pretest

A series of pretests were carried out to select the core product, two extension products and fictitious brands and ads. [Please contact the author for the further details of the findings.]

Stage One: Selecting core product and extension products. Denim clothing was chosen as a core product from which two different brand extensions would be made. It was selected because (1) denim clothing was a familiar product to the college students who would be the participants of the study, and therefore should facilitate their cognitive and affective responses to its advertisements; (2) jeans were rated on a moderate level of product involvement (an index of 99 based on Zaichkowsky's Product Involvement Inventory), thus it was likely to minimize the product involvement effect; and (3) jeans were associated with diverse attributes which made it easy for the experimenter to manipulate different attributes in its advertisements. Three different types of denim clothing, women's jeans, the men's jeans and the denim jacket, were used in the study.

Once denim clothing was chosen as the core product, a list was prepared that included product categories that varied in their shared attributes with the denim clothing. The list was presented to a group of 27 students who were from the same population group but independent from those participating in the main study . Students were first asked to write down all their thoughts, other than specific brand names, for each product in the list. They were then asked to rate the overall similarity of each product to each of the three types of denim clothing, and their subjective knowledge of each product and the three types of denim clothing (i.e., how much they thought they knew each product). The purpose of testing the overall similarity was to maintain a moderate level of similarity in both a comparable extension and a moderately noncomparable extension. Subjective knowledge of each product was measured to separate the comparability between denim clothing and a brand extension from the individuals' knowledge of each product independently.

Sneakers and perfume (or cologne when an extension was made from men's jeans) were chosen as a comparable and a moderately noncomparable extension, respectively, based on the analysis of the product association. Specifically, sneakers as a category shared the same attribute set as the denim clothing although some attributes were more salient in denim clothing than they were in sneakers and vice versa. Perfume/cologne, on the other hand, shared one of its most salient attributes, mystique, with that of denim clothing. But each also had unique attributes that were not comparable with the other. Overall, nine out of ten thoughts associated with denim clothing were also associated with sneakers, while less than half of them were associated with perfume.

A moderate level of similarity was maintained for both sneakers (4.68 on a scale of 1-7) and perfume/cologne (3.77). There was no significant difference in the level of subjective knowledge across the three types of denim clothing and the two extension categories (p>.05). Therefore, the difference in comparability is not from the difference in the participants' independent knowledge of denim clothing and sneakers (or perfume/cologne), but from their relationship to each other.

Stage Two: Selecting ad stimuli and brand names. Six print advertisements (two for each of the three types of denim clothing) were prepared either from unfamiliar magazine advertisements or from pictures that could be disguised as advertisements for denim clothing. It was important to use ads that were not familiar but looked real to the participants because their cognitive and affective responses to the ad were crucial measures in the study. The real brand names and any other written information were replaced with the prepared headlines using a computerized scanner.

Three fictitious brand names were borrowed from a previous study (Hitchon & Churchill, 1992). They were Belas, Komar and Jasil. Three random combinations were prepared between the three brand names and the three types of denim clothing to cancel out any difference that was uniquely associated with a specific brand name. The three combinations were then inserted to each of the six advertisement , thus creating a total of 18 ad stimuli. The ads were then developed into slides.

Procedure

A total of 132 communication major students from a large Midwestern University participated in the study in groups of three to seven. Each group was randomly assigned to one of the three orders of different types of denim clothing. Within each assigned order, each group was again randomly assigned to one of the three brand names. Once the participants were seated, the experimenter explained that they were participating in a part of advertising copy test for a national clothing company. They were told that the real brand names were replaced by fictitious names for confidentiality.

When an introduction was over, a slide of the advertisement corresponding to the assigned order was shown to the participants for 10-15 seconds. They were then asked to list all of their thoughts about the advertised brand. Participants' perception of the brand was then measured with a scale that included both salient and latent attributes of the brand. It was followed by a measure of their overall attitude toward the ad (Aad), attitude toward the brand (Ab), and familiarity with the ad.

The experimenter, then, briefly explained the concept of brand extension by saying, "it refers to a situation in which a company that currently makes a certain product with a brand name 'A' introduces a new product under the same brand name." Ivory shampoo and Honda lawnmower were presented as examples of brand extensions from Ivory soap and Honda automobile, respectively. Once all of the participants said that they understood the concept , they were told that a brand extension, either sneakers or perfume (cologne if the core product was men's jeans), would soon be introduced from the advertised brand of denim clothing. They were then asked to list all of their thoughts at the moment on the specified brand extension. This was followed by a scale-measure of the perceived appropriateness of the brand attributes for the brand extension. The items in the scale were the same as those used to measure their perception of the advertised brand. Next, their overall attitude toward a brand extension (Ab-ext) was measured, followed by their perception of the overall similarity of the brand extension to the advertised brand. The same procedure was repeated for the remaining two types of denim clothing. After subjects completed all three sections, they were asked what they thought the purpose of the study was; if they were aware of the stated purpose while answering the questions; and if they would write down more details if they did. Finally, they were debriefed and dismissed.

Measures

Subjects' attitude toward the advertisement (Aad), was measured with six seven-point semantic differential items that were used in a previous research (Stayman & Batra, 1991). They are 'unpleasant-pleasant,' 'unfavorable-favorable,' 'dislike-like,' 'low quality-high quality,' 'worthless-valuable' and 'disagreeable-agreeable. Since these items were highly correlated with each other (Cronbach's Alpha=.96), the mean rating across six items was used as a single measure of individuals' attitude toward the advertisement. The same items were used to measure their attitude toward the brand (Ab), and attitude toward the brand extension (Ab-ext) .

Accessibility of the brand attributes was measured from the participants' thought listing. Participants' thoughts on each brand extension was differentiated into individual thoughts and coded by two independent judges depending on whether each thought was based on the participant's prior description of the brand (brand-relevant thoughts), or on something else, including the unique attributes of extension category (brand-irrelevant thoughts). The inter-judge agreement was 86%. Disagreement was resolved by discussion, thus all thoughts were coded. Once the coding was completed, the number of brand-relevant thoughts in each subject's response was counted to be a single measure of his or her accessibility of brand attributes.

Perceived appropriateness of the brand attributes for a brand extension was measured by the participants' rating of a brand extension on a scale of six seven-point semantic differential items that represented the brand-salient attributes. Three of the six items in the scale were the salient attributes of the denim clothing that were also salient in the category of sneakers, but less so in the category of perfume/cologne ("intrinsic attributes"). They were: 'not durable-durable,' 'not lasting-lasting,' 'not built well-built well' (for women's jeans ); 'uncomfortable-comfortable,' 'not easy to wear-easy to wear,' 'not relaxing to wear-relaxing to wear' (for men's jeans); 'not warm-warm,' 'not weather-resistant-weather resistant,' 'doesn't protect from the elements-protects from the elements' (for denim jacket). The inter-item reliability coefficients (Cronbach's Alpha) for each scale were .89, .92, and .75, respectively. The other three items in the scale were the salient attributes of the denim clothing that were also salient in the category of perfume/cologne more than in sneakers ("extrinsic attributes"). They are: 'gentle-bold,' 'unconfident-confident,' 'not sexy-sexy' (for women's jeans); 'soft-rugged,' 'weak-tough,' 'feminine-masculine' (for men's jeans); 'not exotic-exotic' 'not mysterious-mysterious' 'timid-bold' (for denim jacket). The inter-item reliability coefficients (Cronbach's Alpha) for each scale were .76, .86., and .76. An average score across the three items in each scale was used as an index of the perceived appropriateness of brand attributes for a brand extension.

TABLE 1

EFFECTS OF AAD AND EXTENSION COMPARABILITY ON THE NUMBER OF BRAND-RELEVANT THOUGHTS

RESULTS

Treatment and Control Check

Order Effect. Because each participant saw three advertisements (one for each of the three types of denim clothing), it was suspected that the order of presentation might have a significant effect on the dependent measures in addition to the effects from the independent variables. Hierarchical regression analyses were carried out to test the main effect of the order and the interaction effect with the independent variables (Aad and extension comparability). The results showed that the main effect of order was significant for one of the dependent measures (perceived appropriateness of the extrinsic attributes). The interaction effect of the order with the independent variables was not significant for any of the dependent measures. Thus, the effect of order was removed first (co-varied out) before testing the effect of the independent variables in the analyses to test hypotheses.

Brand-attribute accessibility in no-ad exposure group. There was no significant difference in the number of brand-relevant thoughts between the two brand extensions among those who had not seen an ad. (X=2.09 for sneakers, X=1.51 for perfume/cologne, p>.05). Thus, it may be said that any significant differences in the accessibility of brand attributes between the two types of extensions among those who had seen an advertisement was not due to an inherent difference between sneakers and perfume/cologne as categories.

Hypotheses Testing

Effect of Aad and Extension Comparability on Brand-Attribute Accessibility. H1a suggested that the more positive consumers' attitude toward the ad (Aad) was, the higher the accessibility of brand attributes was in their perception of a brand extension. A hierarchical regression analysis in Table 1 showed that Aad explained a small, but significant amount of variance in the number of brand-relevant thoughts the participants listed for brand extensions (R²=.04, p<.01). On average, those who had a positive Aad (scored 5 or higher on a scale of 1-7) listed more brand-relevant thoughts (X=3.32) than those who had a negative Aad (scored below 3 on a scale of 1-7) did (X=2.51). Therefore, H1a was supported.

It was also hypothesized (H1b) that the effect of Aad on the brand-attribute accessibility would be stronger for a noncomparable extension than for a comparable extension. The results showed that the interaction between Aad and the extension comparability was not significant (p>.05). That is, a positive Aad increased the number of brand-relevant thoughts to a similar degree for both comparable and noncomparable extensions. Therefore, H1b was not supported.

However, given the significant main effect of the Aad, the author was interested in examining whether the effect might have occurred differently across different levels of Aad (i.e., positive, neutral, negative). This was because the previous research on affect and cognitive organization suggested that the effect between positive and negative affect was not always symmetric. Furthermore, studies have shown that positive affect significantly increased cue-accessibility when the task was ambiguous, but not when it was easy. Thus, the average number of brand-relevant thoughts in each of the three levels of Aad was compared between a comparable extension and a noncomparable extension. The overall trend in Figure 1 showed that the effect of the Aad was mostly created by those who had a positive Aad as opposed to those who had a neutral Aad when a brand extension was made to a noncomparable category (X₂₃=3.41 vs. X₂₂=2.44). [The first number in the subscript refers to the extension category (1=comparable, 2=noncomparable). The second number refers to the level of Aad-core (1=negative, 2=neutral, 3=positive).] Positive Aad little influenced the accessibility of brand attributes when the extension was made to a comparable category (X₁₃=3.23 vs. X₁₂=2.95). On the other hand, negative Aad decreased the accessibility of brand attributes over neutral Aad for a comparable extension (X₁₁=2.28 vs. X₁₂=2.95) while it had little influence for a noncomparable extension (X₂₁=2.73 vs. X₂₂=2.44). Therefore, there were some preliminary indications that (1) the effect of consumers' attitude toward brand advertising might have a nonlinear relationship with the accessibility of the brand attributes, and (2) the pattern of relationship might differ between a comparable and a moderately noncomparable extension.

Underlying process. Given that positive Aad increased the brand-attribute accessibility in consumer perception of a brand extension, an additional analysis was performed to determine if positive Aad increased the number of brand-relevant thoughts in part by increasing the number of thoughts on the advertised brand itself. That is, consumers who liked brand advertising might have had more thoughts on the brand to begin with. They were then probably better able to retrieve those brand-relevant thoughts later when they were asked to describe a brand extension.

FIGURE 1

EFFECT OF AAD AND EXTENSION COMPARABILITY ON THE NUMBER OF BRAND-RELEVANT THOUGHTS

A correlation analysis showed that the number of thoughts on the advertised brand was positively correlated with the number of brand-relevant thoughts on brand extensions (r=.41, p<.001). Aad, however, was not significantly correlated with the number of thoughts on the advertised brand (r=.07, p>.05). As a result, a hierarchical regression analysis showed that each of the two variables, the number of thoughts on the advertised-brand and Aad, had a unique and significant effect on the number of brand-relevant thoughts in brand extensions (Partial corr coeff.=.40, p<.001 for the number of thoughts on core-brand; Partial corr coeff.=.15, p=.05 for Aad). Therefore, Aad had a significant direct effect on the accessibility of the brand-attributes in consumers' perceptions of brand extensions.

Effect of Aad and Extension Comparability on Perceived Appropriateness of Brand Attributes. It was suggested that consumers' attitude toward the brand advertising would have a significant effect on the degree to which they thought the brand attributes were also appropriate for a brand extension. Specifically, it was hypothesized (H2a) that the more positive Aad was, the more consumers would think that the brand attributes were also appropriate for a brand extension. As expected, the more positive Aad was, the higher were the ratings on the perceived appropriateness of the brand attributes. Two separate hierarchical regression analyses (Table 2) showed that the main effect of Aad was overall strong and significant (p<.001) for the brand attributes whether they were category-congruent or category-incongruent. Therefore, H2a was supported.

It was also hypothesized (H2b) that the effect of Aad on the perceived appropriateness of the brand attributes would be more consistent for a noncomparable extension than for a comparable extension. Figure 2 shows the pattern of interaction. For a comparable extension, Aad had a differential influence on the perceived appropriateness of the brand attributes, depending on their congruity with the category concept. It had a greater effect for the category-congruent attributes than it did for the category-incongruent attributes. For a noncomparable extension, on the other hand, Aad showed consistent effects on the perceived appropriateness of the brand attributes regardless of their congruity with the extension category. As a result, hierarchical regressions (Table 2) showed that the interaction between Aad and extension comparability was not significant for the category-congruent attributes (p>.05), but was significant for the category-incongruent attributes (p<.05). Therefore, H2b was supported.

Underlying process. Another issue of interest was the extent to which Aad had a direct effect on the perceived appropriateness of the brand attributes in a brand extension, and/or Aad had an indirect (i.e., mediated) effect through its effect on the brand itself. A separate analysis showed that Aad was positively correlated with the ratings of the brand attributes in the brand itself (r=.45, p<.001). That is, the more positive the participants' Aad was, the more they were likely to agree that the ad-claimed attributes appropriately described the brand. This, in turn, influenced the degree to which they thought that those attributes were also appropriate for the brand extensions. A hierarchical regression analysis was performed to see if the unique effect of Aad on the perceived appropriateness of the brand attributes in a brand extension remained significant after accounting for its effect on the brand itself. The result indicated that the unique effect of Aad became smaller, but remained significant even after accounting for its effect on the advertised brand (Partial corr coeff.=.11, p<.05). Therefore, Aad had both a significant direct effect and an indirect effect (through its influence on the brand) on the perceived appropriateness of the brand attributes in a brand extension.

Nonhypothesized Findings

Effect of Aad and Extension Comparability on Attitude Toward Brand Extension. Although no specific hypotheses were proposed, the effect of Aad on the participants' overall attitude toward the brand extension (Ab-ext) was also measured. The result suggested that Aad had a strong effect on their overall attitude toward the brand extension. The Aad explained more than a-third of the total variance in the overall attitude toward the brand extension (R² change=.37, p<.001). Furthermore, the interaction between Aad and the extension comparability was significant (p<.05). That is, the overall increase occurred to a greater degree for a noncomparable extension than for a comparable extension. The average rating increased from 2.62 (among those who had a negative Aad) to 5.32 (among those who had a positive Aad) for a noncomparable extension, while it increased from 3.17 to 4.82 for a comparable extension.

TABLE 2

EFFECTS OF AAD AND EXTENSION COMPARABILITY ON THE RATING OF PERCEIVED APPROPRIATENESS OF BRAND ATTRIBUTES

FIGURE 2

EFFECT OF AAD AND EXTENSION COMPARABILITY ON THE RATING OF PERCEIVED APPROPRIATENESS OF BRAND ATTRIBUTES

DISCUSSION

Using fictitious brand names and realistic-looking advertisements may have overestimated, to some degree, the effect of advertising over a more stable effect of the brand itself. In addition, subjects were asked about a brand extension shortly after they gave their responses to the advertised brand. In the real world, there most likely will be a longer delay between consumers' exposure to brand advertising and to a brand extension. Thus, their responses to brand advertising may not be as salient in consumers' minds as they were in the current study. On the other hand, consumers are likely to be exposed to brand advertising more frequently in the real world. A previous study suggested that the frequency effect was stronger than the recency effect when there was a long delay between priming and an observation (Higgins, Bargh and Lombardi, 1985). Nevertheless, a future study may consider using a distraction task between participants' exposure to the advertisement and their response to a brand extension.

Within this and other limitations, findings from the current study seem to imply that consumers' attitude toward the brand advertising is important not only for the brand itself, but also for a future brand extension, especially when an extension is planned to a somewhat noncomparable category. Affect associated with the brand advertising shows a significant impact on they way consumers relate a brand extension to the brand, such as accessibility and perceived appropriateness of the brand attributes. Furthermore, the preliminary analyses suggest that the effect tends to occur to a greater degree (or more consistently) for a moderately noncomparable extension than for a comparable extension. These indications are consistent with the findings from previous research on the relationship between affect and cognitive organization. Therefore, consumers' attitude toward brand advertising may influence their perceptions of a brand extension in a similar way that an overall affective state influences individuals' cognitive organization.

REFERENCES

Aaker, David A. and Kevin Lane Keller (1990), "Consumer Evaluations of Brand Extensions," Journal of Marketing, 54 (January), 27-41.

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

Branscombe, Nyla R. and Brian M. Cohen (1991), "Motivation and Complexity Levels as Determinants of Heuristic Use in Social Judgment," in Emotion & Social Judgment, ed., Joseph Forgas, New York: Pergamon Press, 145-160.

Cialdini, R. B., B. Darby and J. Vincent (1973), "Transgression and Altruism: A Case of Hedonism," Journal of Experimental Social Psychology, 9, 502-516.

Cohen, Joel B. and Charles S. Areni (1990), "Affect and Consumer Behavior," in Handbook of Consumer Theory and Research, eds. Harold Kassarjian and Thomas S. Robertson, Englewood Cliffs, NJ: Prentice Hall, 188-240.

Edell, Julie E. and Marian Chapman Burke (1992), "The Impact of Memorability of Ad-Induced Feelings: Implications for Brand Equity," in Advertising and Building Strong Brands, eds. David A. Aaker and Alexander L. Biel, in press.

Forgas, Joseph P. (1991), "Affect and Social Judgments: An Introductory Review," in Emotion and Social Judgments, (ed.), Joseph P. Forgas, New York: Pergamon Press, 3-29.

Forgas, Joseph P. and Gordon. H. Bower (1987), "Mood Effects on Person-Perception Judgment," Journal of Personality and Social Psychology, 53, 53-60.

Hitchon, Jacquelyn and Gilbert Churchill (1992), "To Be or What To Be: An Empirical Investigation of Metaphor Effects in Advertising Commercials," unpublished manuscript, University of Wisconsin-Madison.

Isen, Alice M. (1984), "Toward Understanding the Role of Affect in Cognition," in Handbook of Social Cognition, eds., Robert S. Wyer, Jr. and Thomas K. Srull, Hillsdale, NJ: Erlbaum, 179-236.

Isen, Alice M. (1987), "Positive Affect, Cognitive Process, and Social Behavior," in Advances in Experimental Social Psychology, ed., Leonard Berkowitz, New York: Academic Press. 20, 203-253.

Isen, Alice M. and Kimberly A. Daubman (1984), "The Influence of Affect on Categorization," Journal of Personality and Social Psychology, 47 (6), 1206-1217.

Johnson, Michael D. (1984), "Consumer Choice Strategies for Comparing Noncomparable Alternatives," Journal of Consumer Research, 11 (December), 741-753.

Keller, Kevin Lane (1987), "Memory Factors in Advertising: The Effects of Advertising Retrieval Cues on Brand Evaluation," Journal of Consumer Research, 14 (December), 316-334.

Kunda, Z. (1987), "Motivated Inferences: Self-Serving Generation and Evaluation of Causal Theories, Journal of Personality and Social Psychology, 53, 636-647.

Murray, N. H., E. R. Hirt Sujan and Mita Sujan (1990), "The Influence of Mood on Categorization: A Cognitive Flexibility Hypothesis," Journal of Personality and Social Psychology, 59, 411-425.

Park, C. Whan, Sandra Milberg, and Robert Lawson (1991), "evaluation of Brand Extensions: The Role of Product Feature Similarity and Brand Concept Consistency," Journal of Consumer Research, 18 (September), 185-193.

Schwarz, Norbert and Herbert Bless (1991), "Happy and Mindless, But Sad and Smart? The Impact of Affective States on Analytic Reasoning," in Emotion and Social Judgment, ed. Joseph Forgas, New York: Pergamon Press, 55-71.

Schwarz, Norbert and Gerald L. Clore (1983), "Mood, Misattribution, and Judgment of Well-Being: Informative and Directive Functions of Affective States," Journal of Personality and Social Psychology, 45, 513-523.

Stayman, Douglas M. and Rajeev Batra (1991), "Encoding and Retrieval of Ad Affect in Memory," Journal of Marketing Research, 28 (May), 232-239.

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