Goal-Related Consumption and Extension Advertising: the Impact on Memory and Consumption

Brian Wansink, Dartmouth College
Michael L. Ray, Stanford University
ABSTRACT - Advertising which encourages a person to use or to consume a familiar product in a new situation is referred to here as extension advertising. This paper examines some of the delayed effects of extension advertising. In particular, it shows that these extension ads are more memorable when they explicitly make comparisons with either other products or with other situations. It also shows that regardless of the impact that these ads might have on a meal planner, their impact on actual consumption will be mediated by the attitudes of other family members.
[ to cite ]:
Brian Wansink and Michael L. Ray (1992) ,"Goal-Related Consumption and Extension Advertising: the Impact on Memory and Consumption", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 806-812.

Advances in Consumer Research Volume 19, 1992      Pages 806-812

GOAL-RELATED CONSUMPTION AND EXTENSION ADVERTISING: THE IMPACT ON MEMORY AND CONSUMPTION

Brian Wansink, Dartmouth College

Michael L. Ray, Stanford University

ABSTRACT -

Advertising which encourages a person to use or to consume a familiar product in a new situation is referred to here as extension advertising. This paper examines some of the delayed effects of extension advertising. In particular, it shows that these extension ads are more memorable when they explicitly make comparisons with either other products or with other situations. It also shows that regardless of the impact that these ads might have on a meal planner, their impact on actual consumption will be mediated by the attitudes of other family members.

Advertising which encourages a relatively new or different use of a product will be referred to as extension advertising. Arm and Hammer Baking Soda launched a very effective extension advertising campaign when it extended the use of baking soda from a baking product to a refrigerator deodorizing product. Similarly, the Florida Citrus Commission used an extension advertising campaign when it told consumers that orange juice was "not just for breakfast anymore," and could be extended into the afternoon, and after exercising. In general, extension advertising includes any advertising which encourages a person to consume a product in a situation in which it is generally not consumed, and extension advertising is defined with respect to a particular segment of consumers who do not use the product in this specific situation, or who do so very infrequently and only as a very secondary use of the product (Wansink and Ray, 1992).

In general, we commonly see extension ads which either compare or frame the target product with another product or with its use in another situation. Such framed ads can emphasize a frame of reference which encourages consumers to think of using the product either 1) with respect to the other products one conventionally consumes in that situation (product frame), or 2) with respect to the other situations in which one conventionally consumes this brand (situation frame). Though extension ads (such as the Arm and Hammer ads) need not contain any overt comparison or frame of reference (they can be "unframed"), this research shows that both situation frames and product frames will be more effective than unframed ads in causing one to remember an ad for eating a particular product in a particular situation. In addition, the research will argue that the consumption of that product will be influenced more by a meal planners perceptions of family preferences than by his or her own preferences.

BACKGROUND: CONSUMPTION FRAMING

A consumption frame refers to either the point of comparison used in an extension ad or to the point of comparison used by a consumer when processing information about an extension. Even though consumption framing can be encouraged through the structure and content of an ad, it only has an impact when consumers perceive the frame.

An ad with a product frame is one that "positions" or compares the target brand with selected attributes of another product that is already favorably associated with that situation. For instance, such an ad (or commercial) might begin with a woman serving the conventional product (one that is strongly associated with the situation) in the target situation. She might say that benefit1 and benefit2 make it a reasonable choice for the situation. The scene would then change and she would be serving the target brand in the target situation. She says that just as benefit1 and benefit2 made the conventional product reasonable in this target situation, they also make the target brand reasonable for that situation. For instance, an ad that encourages consumers to eat Campbell's Soup with their dinner might compare some attributes of Campbell's Soup (e.g., quick and nutritious) with similar attributes of a product they conventionally consume with dinner (e.g., lettuce salad).

Whereas a product frame draws similarities between the target product and a conventional product, a situation frame draws similarities between consuming the target product in a new situation with consuming it in a more conventional situation. The basic situation-framed ad might begin with a woman serving the target brand in the situation in which it is most conventionally served. She might say that benefit1 and benefit2 make it an reasonable choice for the situation. The scene would then change and she would be serving it in the target situation. She says that just as benefit1 and benefit2 make the brand reasonable in this conventional situation, they also make it a reasonable choice for this target situation. For instance, an ad which encourages consumers to eat Campbell's Soup with their dinner might compare how eating it at dinner is similar to eating it in a more conventional situation, such as with lunch. By doing this, a situation frame emphasizes an additional situation in which the product can be consumed.

In contrast to these, an unframed ad would draw no simularities with other products or other situations. The basic unframed ad might begin with a woman serving the target brand in the target situation. She would simply say that benefit1 and benefit2 make it an reasonable choice for the situation. Essentially the arguments she gives (i.e., the attributes she mentions) for eating the target product in the target situation are identical to those she would have given in either the product frame or the situation frame. The only differences are the points of comparison that are suggested (or that are not suggested).

HYPOTHESES

Consumption Framing and Memory

One's actual consumption of the target product in the target situation not only depends on how the extension ad influences one's attitude toward the extension (Aext) but also on whether it affects one's ability to recall or evoke the product in the appropriate consumption situation. Indeed, the cued recall of the product in the target situation is a necessary condition of consumption.

Though product frames and situation frames encourage different types of comparisons and different types of thinking, they should both be more effective than unframed ads in increasing the number (and strength) of associations that are made between the target product and the target situation. Whereas a product frame should increase the number of associations made between the target brand and a product that is already associated with the situation, a situation frame should increase the number of associations made between the new situation and more conventional situations in which the brand is consumed.

Since one's ability to recall or to evoke a brand is dependent on the number and on the strength of associations between the brand and the situation (Nedungadi 1990), the more associations that an ad can encourage a person to make between the target product and the target situation, the more likely one will recall that target brand when given that situational cue at a later date (Keller 1991). Therefore . . .

H1: When compared with an unframed ad, framed ads (i.e., product frames or situation frames) will generate greater recall when given a situational cue.

Extension Advertising and Consumption

Though this study focuses solely on the long-term impact of extension advertising, such advertising does have immediate impacts on one's attitude toward the extension (Aext) and on beliefs (bi) about the product and on evaluations (ei) as to whether these beliefs are relevant in a particular situation. As is consistent with Fishbein and Azjen (1975), Aext can be approximated as _biei. It has been argued by Azjen and Fishbein (1980) that any lack of correspondence between attitudes and behavior (i.e., Aext and actual consumption in that situation) can often be attributed to external factors, such as to what others (such as family members) think about the behavior (behavioral norms -- bni) and the degree to which abiding by these people's thoughts is important to a person (motivation to comply -- mci). This less-than-perfect relationship between one's attitude and one's actual behavior is particularly vivid when meal planning is involved (Rudell 1979; Krishnamurthi 1983). As a result, we see meal planners who love liver and onions, but who would never serve this at their home because no one else would eat it.

Nevertheless, research has suggested that a researcher may be able to more effectively predict consumption behavior by taking these attitudes of relevant others (such as family members) into account (Pantzar 1988). Ajzen and Fishbein (1980) viewed this component as the perceived social norm (SN) and modelled it as the summed cross-product of behavior norms (bn) and one's motivation to comply (mc) with these norms (_bnimci). The social norm of serving Campbell's Soup for breakfast would consist of what a meal planner believed his or her family would think of it (bni) as well as the degree to which he or she cares about what family members think (mci). When these norms are specific to a certain extension situation (e.g., serving soup for breakfast), they are modelled as _bnextmcext.

Though Axelson, Brinberg, and Allen (1986) found that the correlations between SN and behavioral intentions were very low for both college students (r=.10) and for families (r=.23), SN may be of predictive value when combined with attitude information. In this instance, we should find that the social norms of a family can profoundly mediate the impact of the meal planner's Aext on actual consumption. Regardless of how effective an extension ad is in changing the Aext of a meal planner, if he or she does not think family members will like it, he or she will be unlikely to serve it. In summary,

H2: The perceived social norms within one's family will mediate the relationship between a meal planner's Aext and actual consumption.

METHODOLOGY

The goals of the main study were 1) to determine how various frames impact one's memory, and 2) to determine how subsequent consumption is mediated by family considerations.

Subjects and Design

Adult subjects were recruited though PTA groups, and six dollars were donated to the respective organization for each of the 219 group members who participated. The age of the subjects ranged from 26 to 67 with 83 percent being between the ages of 30 and 45. The majority (86 percent) of these subjects were not employed outside the home. Their educational backgrounds were heterogeneous.

The basic experiment was a 4 x 3 x 2 between-subjects design with four levels of framing (product frame, situation frame, no frame, and a "no ad" control condition), three brands (Campbell's Soup, Jell-O Brand Gelatin, and Ocean Spray Cranberry Sauce), and two levels of appropriateness (appropriate extensions and less appropriate extensions). The products and their relative appropriate and less appropriate extensions were determined in a prestudy. In essence, the last two factors (brands and appropriateness) were included for generalizability and will not be factors which will be individually analyzed.

Procedure

Upon beginning the experiment, the subjects were told to take alternate seats, and they were given a closed packet of materials which contained a cover sheet of instructions and a number of consecutively labeled booklets. The subjects were told they were going to be comparing how understandable transcripts of commercials were to storyboards of commercials. They were then asked to read through a transcript for the target brand in a natural way, write their thoughts and feelings about it, answer the questions which followed it, read an unrelated storyboard, and to answer more questions. The objective of the cover story was to insure that subjects would process and comprehend information in the ad without unduly focusing on directly evaluating the product. They were last asked demographic questions and questions related to the manipulation checks. Three months after their involvement in the study, they were called and asked about their memory of the target product and their consumption of it.

Measures

After writing down their cognitive responses, subjects were asked a couple of questions which served as a basic set of manipulation checks regarding the frame. These questions were quantitative processing measures which used a Likert scale to rate ("Strongly Disagree" = 1; "Strongly Agree" = 7) the degree to which subjects focused on the brand or on the situation.

Following this, a measure of Aext was taken by asking subjects to indicate (on a seven-point scale) whether a particular extension was "Bad-Good," "Unappealing-Appealing," "Inappropriate-Appropriate," and "Unreasonable-Reasonable." The Cronbach's alpha (.921), was high enough to enable Aext to be analyzed as the average of the four items (Nunnally 1967). Make this consistent with past research on multi-attribute models).

All subjects were then asked how many times in the past year they consumed the basic product category and the percentage of times they consumed the target brand, and they were asked the degree to which they conformed to the consumption attitudes of their children and spouse ("need to comply") and how their children and spouse would feel about eating the target brand in this extended situation ("social norm"). As has been recommended (Fishbein and Ajzen 1980), these measures used a seven-point Likert scale ("Strongly Disagree" = 1; "Strongly Agree" = 7).

To determine a subject's memory of the brand and their reported consumption of it, subjects were telephoned roughly three months after they were involved in the experiment. They were given the appropriate situation cue for the ad they saw (breakfast, snack, dinner) and were asked if they could recall the brand in the transcript. They were then asked how many times (since the experiment) they consumed the target brand in the relevant extension situation. Though such measures are not without the typical flaws and biases associated with self-report measures of consumption (Jones 1985), they were taken with the hope of adding richness to the study and adding credibility to the results.

Stimuli

Each subject saw a professionally edited ad (in the form of a transcript) for one of three brands. In total, there were 18 different ads: There were three different frames (product-, situation-, and un-framed ads) for each of the three brands (Campbell's Soup, Jell-O Brand Gelatin, and Ocean Spray Cranberry Sauce) and for each of two extensions (appropriate and less appropriate). Though this study only examines the difference between the framing conditions, by using multiple brands and a wide range of extension appropriateness, we are able to strengthen the generalizability of this study.

The eighteen hypothetical ads mentioned above were presented in the form of transcripts and each contained the two claims that were determined for each brand through the preliminary studies. The initial drafts were professionally edited by a copywriter who had significant experience with consumer products, and they were then subsequently re-edited to eliminate all irrelevant variation between conditions. These transcripts used either a product frame, a situation frame, or no frame.

The basic un-framed transcript begins with a woman serving the target brand in the target situation. She mentions why benefit1 and benefit2 make the brand a reasonable choice for the situation. The commercial then ends with a slogan containing 1) the target brand, 2) this target situation, and 3) the two benefits.

The basic product-framed transcript begins with a woman serving the conventional product (one that is strongly associated with the situation) in the target situation. She mentions why benefit1 and benefit2 make it a reasonable choice for the situation. The scene then changes and she is now serving the target brand in the target situation. She says that just as benefit1 and benefit2 made the conventional product reasonable in this target situation, they also make the target brand reasonable for that situation. The commercial then ends with a slogan containing 1) the target brand, 2) this target situation, 3) the two benefits, and 4) the conventional product.

The basic situation-framed transcript begins with a woman serving the target brand in the situation in which it is most conventionally served. She mentions why benefit1 and benefit2 make it an reasonable choice for the situation. The scene then changes and she is now serving it in the target situation. She says that just as benefit1 and benefit2 make the brand reasonable in its conventional situation, they also make it a reasonable choice for this target situation. The commercial then ends with a slogan containing 1) the target brand, 2) this target situation, 3) the two benefits, and 4) the conventional consumption situation.

RESULTS

Manipulation Checks and Overview of Analyses

Recall that this study's focus is to determine the impact that extension advertising has on consumers who are already "users" of a brand. A "user" was liberally defined as someone who had consumed the target brand at least once in the past year. This measure of usage also turned out to be highly related to conservative measures of brand loyalty (Jacoby and Chestnut 1978), since the typical individual in this study consumed the target brand 84 percent of the time he or she used a product from that particular product category. The subjects' prior consumption of each brand was similar across all the conditions.

The manipulation of the product frame and the situation frame appeared to be successful. Quantitative measures of processing ("I compared serving . . .") were taken on 7-point scales. They indicated that the subjects seeing situation frames "compared" the conventional situation and the target situation to a greater degree than did those subjects who saw unframed ads [F(1,98) = 14.6, p < .01] or to those subjects who saw product frame ads [F(1,96) = 39.8, p < .01]. Similarly, subjects seeing product frames compared the target brand and the conventional brand to a greater degree than did those subjects who saw unframed ads [F(1,97) = 9.3, p < .05] or to those subjects who saw situation frame ads [F(1,98) = 12.4, p < .01].

Though the hypotheses were assumed to be generalizable across all three brands and across both levels of "appropriateness," it was necessary to determine if the results for each of the brands were indeed similar enough to justify this. In general, the three brands involved in the study all exhibited similar patterns with regard to how subjects reacted toward them in both the frame and appropriateness conditions. A 3-way ANOVA (Brand by Frame by Appropriateness) was conducted on a number of key dependent variables, and the results indicated that there were mean-level differences between the brands but no significant interactions. The results indicated that there were also mean-level differences between the two levels of appropriateness but that there were no interactions. Because the basic patterns of the data were similar, the analyses were combined and each brand and each level of appropriateness was represented by a dummy variable in the analyses in order to account for their mean-level difference in response. Wansink (1990) provides a detailed reporting of this procedure.

Memory and Consumption

Subjects were phoned approximately three months after they participated in the experiment, and they were asked about their recall of the target brand and how many times they consumed the target brand in the relevant extension situation. Of those 219 people originally participating in the study, 181 agreed to provide information regarding this consumption.

There were two objectives when collecting the delay data: 1) To determine if framed ads promote cued brand recall better than unframed ads (H1), and 2) to determine the extent to which actual consumption is influenced by the social norms of one's family (H2).

When given the relevant situation cue, 73 percent of those people who saw product frames were able to recall the target brand. This was significantly different than the 42 percent recall of the people who saw unframed ads [F(1,84)=7.34; p<.01] and directionally different from the 58 percent of those who saw situation frames. (See Figure 1). A linear contrast showed that framed ads (situation frames and product frames) were more effective than the unframed ones (p<.05). Though this finding was hypothesized (H1), it was inconsistent with the cognitive responses of the subjects, for framed ads did not generate more cognitive responses than unframed ads. It may be, however, that the thoughts that were generated by framed ads stimulated associations that were stronger or more relevant than those stimulated by unframed ads.

When examining the general impact that these ads had on actual consumption, it was hypothesized that the correlation between Aext and reported consumption (r=.24) would be greatly moderated by the social norms of one's family (H2).

Measures of social norms (SN) were obtained by calculating the summed cross-product of behavior norms (bni) and one's motivation to comply with these norms (mci) (Ajzen and Fishbein 1980). The norms consisted of a two-item score which asked subjects whether their spouse would approve of consuming the product in this situation (1="Strongly Disagree"; 7="Strongly Agree") and whether their children would approve.

Given this measure of SN, a regression of SN, Aext, and one's prior consumption frequency of the brand was conducted on actual extended consumption. Having included both appropriate and less appropriate extensions in the study should raise the variance of social norms, thus allowing us to effectively examine when SN are significant and when they are not. To test this, one regression was conducted using only the appropriate extensions, and a second regression was conducted using only inappropriate extensions. The results confirmed H2, for these regressions indicated that SN was insignificant when the extension was appropriate but significant when it was inappropriate (t=2.35; p<.05).

Regardless of whether an extension ad influences Aext, it still had a minimal impact on consumption if a meal planner thought his or her family would dislike it. Though changing the Aext of the meal planner is an important precondition to changing consumption, this analysis shows it is not sufficient, for one must also consider the Aext of their family.

GENERAL DISCUSSION

Overview and Implications

Recall that people who saw product frames were more likely to recall the brand when provided a situational cue three months later (Table 1).

FIGURE 1

IMPACT OF CONSUMPTION FRAMES ON CUED RECALL (3 MONTH DELAY)

Specifically, when cued with the extension situation (corresponding to the transcript they had seen three months earlier), 73 percent of the subjects recalled the target product. As mentioned earlier, this was significantly different than those who saw unframed ads (42 percent) and directionally different from those who saw situation frames (58 percent), and it provides evidence for the contention that various comparisons can have a strategic impact if they serve to influence a certain cognitive process, such as memory.

Regardless of the impact that an extension ad has on a meal planner's memory, this research shows that such an impact need not affect subsequent consumption . . . especially when a meal planner is concerned about what family members might think. Such a finding is of important managerial interest since it indicates that the most memorable commercial might not be the most effective commerical if it is targeted only toward the "decision maker" or meal planner and is not also targeted at other family members. As a result, one might want to use a "pull strategy" by advertising the extension in a clever way to children, thereby encouraging them to remember it and to request it. Conversely, an advertiser might also try to provide the proper "excuse" the meal planner might need to overcome any objections family members might have.

Limitations and Future Research

Great care was taken to make the ad transcripts as comparable as possible (except on the relevant manipulations). The subjects rated all the transcripts as being equally clear and as requiring an equal amount of concentration. The reading of these transcripts, however, was performed in a high involvement context. Nevertheless, if a person is highly involved with a stimulus, there appears to be no reason why the same principles operating here would not work in a television campaign, a print campaign, or even on the back of a brand's package. In a low involvement scenario, however, it may be that a product frame can successfully associate the brand with the largest number of situational cues, thus providing positive conditioning (with a high number of exposures) between the brand and the situation (Petty and Cacioppo 1981).

The focus of this study was on brand loyal consumers. Extension ads, however, are likely to have little or no impact on those individuals who are not interested in the product category, or who are brand loyal to another product. Nevertheless, extension ads may have an impact on nonusers who have not yet been introduced to the product category, and these ads may have a differential impact on heavy and light users. It may be that heavy users are more predisposed toward extensions and may thus respond differently to frames because they would be less likely to counterargue against a product frame which suggests a substitution (Wansink and Ray 1992).

SUMMARY

For mature brands which are the market leaders in their category, increasing category volume is likely to be a much more cost-effective strategy for building sales than would be the strategy of increasing market share. Extension advertising provides a viable way to increase category volume since it encourages a person to use or to consume a familiar product in a new situation, referred to here as extension advertising.

This research suggests that the different comparisons made in such ads form or strengthen the associations one has between a product and a situation. In particular, this paper shows that these extension ads are more memorable when they explicitly make comparisons with either other products that are currently used in the target situation, or with other situations in which the target brand is used. In addition, however, this research also shows that regardless of the impact that these ads have on a meal planner's memory for an ad, their impact on actual consumption will be mediated by his or her perception of what what other family members would think.

REFERENCES

Ajen, Icek and Martin Fishbein (1980), Understanding Attitudes and Predicting Social Behavior, Englewood Cliffs, NJ: Prentice-Hall, Inc.

Axelson, Marta L., David Brinberg, and Candace Allen (1986), "Robustness of Fishbein's Theory of Reasoned Action in Predicting Food-Related Behavior," Home Economics Research Journal, Volume 15, 52-56.

Batra, Rajeev and Michael L. Ray (1986), "Situational Effects of Advertising Repetition: The Moderating Influence of Motivation, Ability, and Opportunity to Respond," Journal of Consumer Research, Volume 12, 432-445.

Belk, Russell W. (1974), "An Exploratory Assessment of Situational Effects in Buyer Behavior," Journal of Marketing Research, Volume 11, 156-163.

Belk, Russell W. (1975a), "Situational Variables and Consumer Behavior," Journal of Consumer Research, 2, 157-164.

Bettman, James R. and Mita Sujan (1987), "Effects of Framing on Evaluation of Comparable and Noncomparable Alternatives by Expert and Novice Consumers," Journal of Consumer Research, Volume 14, 141-154.

Burk, Marguerite C. and Eleanor M. Pao (1976), "Methodology for Large-Scale Surveys of Household & Individual Diets," Home Economics Research Report #40. Washington, D.C.: USDA (November)

Canary, Daniel J. and David R. Seibold (1984), Attitudes and Behaviors: An Annotated Bibliography, New York: Praeger Publishers.

Day, George S., Allen D. Shocker, and Rajendra K. Srivastava (1979), "Customer-Oriented Appropches to Identifying Product-Markets," Journal of Marketing, Volume 46, 56-64.

Fischhoff, Baruch (1983), "Predicting Frames," Journal of Experimental Psychology: Learning, Memory, and Cognition, Volume 9, 103-116.

Fishbein, Martin and Icek Ajzen (1975), Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, Reading, MA: Addison-Wesley.

Jacoby, Jacob and Robert W. Chestnut (1978), Brand Loyalty: Measurement and Management, New York: John Wiley and Sons.

Jones, Russell A. (1985), Research Methods in the Social and Behavioral Sciences, Sunderland, MA: Sinauer Associates.

Keller, Kevin Lane (1987), "Memory Factors in Advertising: The Effect of Advertising Retrieval Cues on Brand Evaluations," Journal of Consumer Research, Volume 14, (December), 316-33.

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

Maheswaran, Durairaj and Joan Meyers-Levy (1990), "The Influence of Message Framing and Issue Involvement," Journal of Consumer Research, Volume 27, 361-376.

Marketing Science Institute (1983), "Beyond Recall," Cambridge, MA: Marketing Science Institute.

Nedungadi, Prakash (1990), "Recall and Consumer Consideration Sets: Influencing Choice Without Altering Brand Evaluations," Journal of Consumer Research, December, 263-276.

Nunnally, Jum C. (1967), Psychometric Theory, New York: McGraw-Hill.

Pantzar, Mika (1988), "The Household and Changing Traffic Signs of the Modern Dining Table," Journal of Consumer Behavior and Home Economics, Volume 6, 225-236.

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

Raj, S.P. (1982) "The Effects of Advertising on High and Low Loyalty Consumer Segments," Journal of Consumer Research, Volume 9, 77-89.

Ratneshwar, S. and Allen D. Shocker (1991), "Substitution in Use and the Role of Usage Context in Product Category Structures," Journal of Marketing Research, (forthcoming).

Rossiter, John R. and Larry Percy (1987), Advertising and Promotion Managment, New York: McGraw-Hill Book Company.

Rudell, Fredrica (1979), Consumer Food Selection and Nutrition Information, New York: Praeger.

Krishnamurthi, Lakshman (1983), "The Salience of Relevant Others and Its Effect on Individual and Joint Preferences: An Experimental Investigation." Journal of Consumer Research, Volume 10, 62-72.

Sujan, Mita (1985), "Consumer Knowledge: Effects on Evaluation Strategies Mediating Consumer Judgments," Journal of Consumer Research, Volume 12, 31-46.

Sujan, Mita and Christine Dekleva (1987), "Product Categorization and Inference Making: Some Implications for Comparative Advertising," Journal of Consumer Research, Volume 14, 372-378.

Wansink, Brian (1990), "Consumption Framing and Extension Advertising," unpublished dissertation, Graduate School of Business, Stanford University, Stanford, CA 95305.

Wansink, Brian and Michael L. Ray (1992), "The Impact of Extension Advertising on Brand Equity." In David Aaker and Alex Biel (Editors) Advertising and Building Strong Brands, Cambridge, MA: Lexington.

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