The Effects of Strengthening Category-Brand Associations on Consideration Set Composition and Purchase Intent in Memory-Based Choice

Steven S. Posavac, University of Rochester
David M. Sanbonmatsu, University of Utah
Maria L. Cronley, University of Cincinnati
Frank R. Kardes, University of Cincinnati
ABSTRACT - An experiment was conducted to explore the effects of strengthening the association between particular brands and a superordinate choice category on the likelihood of those brands being included in the consideration set, and chosen in a memory-based choice context. Results showed that a brand was more likely to be present in the consideration set, and indicated as an intended choice, if the association between the brand and the choice category was strengthened vs. not strengthened. Attitudinal data suggest that the positive effects of a brand being strongly associated with the choice category operate independently of attitude toward the brand.
[ to cite ]:
Steven S. Posavac, David M. Sanbonmatsu, Maria L. Cronley, and Frank R. Kardes (2001) ,"The Effects of Strengthening Category-Brand Associations on Consideration Set Composition and Purchase Intent in Memory-Based Choice", in NA - Advances in Consumer Research Volume 28, eds. Mary C. Gilly and Joan Meyers-Levy, Valdosta, GA : Association for Consumer Research, Pages: 186-189.

Advances in Consumer Research Volume 28, 2001     Pages 186-189

THE EFFECTS OF STRENGTHENING CATEGORY-BRAND ASSOCIATIONS ON CONSIDERATION SET COMPOSITION AND PURCHASE INTENT IN MEMORY-BASED CHOICE

Steven S. Posavac, University of Rochester

David M. Sanbonmatsu, University of Utah

Maria L. Cronley, University of Cincinnati

Frank R. Kardes, University of Cincinnati

ABSTRACT -

An experiment was conducted to explore the effects of strengthening the association between particular brands and a superordinate choice category on the likelihood of those brands being included in the consideration set, and chosen in a memory-based choice context. Results showed that a brand was more likely to be present in the consideration set, and indicated as an intended choice, if the association between the brand and the choice category was strengthened vs. not strengthened. Attitudinal data suggest that the positive effects of a brand being strongly associated with the choice category operate independently of attitude toward the brand.

INTRODUCTION

In recent years researchers have become increasingly interested in the processes by which choice options come to receive consideration, and the consequences of consideration effects in decision-making. Some researchers have developed models of the consideration process itself (Ratneshwar and Shocker 1991; Kardes, Kalyanaram, Chandrashekaren and Dornoff 1993; Payne, Bettman and Johnson 1993). Kardes et al., for example, reported the advantages enjoyed by a pioneering brand as the universal set (all brands that may satisfy a consumer need), is pared to the retrieval set (the set of brands a consumer is aware of at the time of decision making), which in turn is pared to the consideration set (set of brands that receive careful consideration prior to choice), and ultimately to choice. Other research has focused on determinants of the composition of the retrieval set (Posavac, Sanbonmatsu and Fazio, 1997) and the consideration set (Lehman and Pan 1994; Shapiro, MacInnis and Heckler 1997). Implicitly, most decision making research in marketing and psychology concerns how individuals make decisions from a consideration set. That is, research on attitudes nd persuasion, multiattribute studies, and so on, typically explore the determinants of choice from a specified set of alternatives.

The importance of each stage of the consideration process may depend on the level of specification of alternatives in the decision context (Nedungadi 1990). In stimulus-based choice, alternatives are specified in the decision context. For example, the decision of which soft drink to purchase at the grocery store is a stimulus-based choice because all of the options are clearly displayed on the grocer’s shelves. Memory-based choice contexts, in contrast, do not contain specified alternatives. Trying to decide on a restaurant after having started the car is a memory-based choice because the recognition that one is hungry suggests only that a restaurant should be sought, but does not specify which alternatives should be considered.

The selection of a particular brand of soft drink, and other stimulus-based consumer choices in which alternatives are highly salient in context, requires the assessment of the options that are present and the choice of a preferred brand. Choosing a restaurant for lunch, and other memory-based decisions, requires the additional step of generating alternatives from memory prior to assessment of the alternatives, and subsequent choice. Figure 1 presents the essential difference between choice contexts that contain vs. do not contain specified alternatives. In memory-based choice, alternatives must be generated from memory prior to the assessment of alternatives and decision making.

When alternatives are not specified in the choice context the likelihood of a particular brand being chosen has been shown to depend on how strongly associated the brand is with the superordinate choice category (Posavac et al. 1997). To the extent that a brand is strongly associated with the choice category, it is likely to be generated upon consideration of the category, and to be present in consumers’ consideration sets. Of course, in memory-based choice only those options that are recalled from memory and identified as options may ultimately be chosen. In stimulus-based choice the category-brand association is less relevant because options are present at the time of decision-making, and accordingly do not need to be generated from memory.

Although prior research has shown that brands strongly associated with the choice category are at an advantage in memory-based choice, this research is correlational, and to date no studies have explored the utility of manipulating the category-brand association in the promotion of particular brands. That is, prior research has related measured associative strength to decision making, but has not manipulated the associative strength of the category-brand association for particular brands. In this study we sought to provide direct evidence that manipulating associative strength may influence consumers’ decision making. More specifically, we explored the possibility that strengthening the association between brands and the choice category would increase the likelihood of those brands being included in the consideration set, and ultimately being chosen, when decision making is memory based.

The effects of manipulating the strength of brand-category associations on choice may operate independently of attitudes toward the brand. Prior research has established reliable choice effects that result from differential likelihood of competing brands being considered prior to choice (Nedungadi 1990). In the present research, we expected that a brand would be at an advantage to the extent that it was manipulated to be strongly associated with the category, but that such a manipulation would not have consequences for attitudes.

Overview of the Research

An experiment was conducted to explore the potential of strengthening the association between particular brands and the choice category on the likelihood that those brands would be present in consumers’ consideration sets, and ultimately chosen. The experiment took place in two sessions, which were separated by a week. In the first experimental session participants engaged in a rehearsal task that strengthened the association between some (rehearsal group), but not other (control group), alternatives and the superordinate choice category to which the alternatives belonged. In the second session participants were asked to indicate choice intentions. They were also asked to report the other options that they seriously considered prior to indicating their choice intentions (i.e., options in their consideration sets in addition to their choice). Choices were always completely memory-based. Specifically, participants indicated their choice intentions, as well as the other alternatives in their consideration sets, on a blank sheet of paper that did not contain the names of any choice options.

FIGURE 1

TASKS REQUISITE TO CHOICE IN STIMULUS-BASED AND MEMORY-BASED DECISION CONTEXTS

We predicted that alternatives that were manipulated to be strongly associated with the choice category (i.e., alternatives in the rehearsal group) would be more likely to be generated from memory than alternatives not rehearsed (i.e., those in the control group). Accordingly, we anticipated that rehearsal group alternatives would be more likely than control group alternatives to be present in participants’ consideration sets. Moreover, we also expected that alternatives in the rehearsal group would be more likely to be chosen than options in the control group. This reasoning suggests the following two hypotheses;

H1: Rehearsal group brands will be more likely than control group brands to be present in participants’ consideration sets.

H2: Rehearsal group brands will be more likely to be chosen than control group brands.

Consistent with reasoning delineated in Nedungadi (1990), Kardes et al. (1993), and Posavac et al. (1997) we expected that these consideration processes would operate independently of attitudes. Accordingly, we predicted that the rehearsal task would not influence attitudes towards the alternatives.

H3: The rehearsal manipulation will not affect attitudes. That is, average attitude toward rehearsal brands will not be different than average attitude toward brands in the control group.

METHOD

Participants

106 university students participated in the experiment in exchange for extra course credit. Students participated individually.

Procedure

First experimental session. Participants recruited for an experiment described as an investigation of cognitive skills were asked to perform several cognitive tasks. After engaging in several distractor activities, participants engaged in a categorization task designed to strengthen the association between some brands and the superordinate choice category to which they belonged (see Posavac et al. 1997 for a similar procedure). Participants were given a list of objects that each belonged to one of four categories, and were asked to place a letter next to each object according to its category membership. For example, participants were instructed to place an "s" next to each object on the list that was a brand of athletic shoe. In an earlier pilot study we identified well-known examples of local Chinese restaurants, local auto service shops that provided oil changes, local pizza restaurants that delivered, as well as shoe brands. Equal numbers of exemplars within each category were randomly assigned to either Group A or Group B. Participants were randomly assigned to rehearse the category membership of exemplars belonging to either Group A or B. Thus, each participant rehearsed the category membership of half of the alternatives within each category. Exemplars appeared six times on the lists. After completing the rehearsal task, participants scheduled a second experimental session and were excused.

Second experimental session. Upon arrival to the second experimental session, participants completed three distractor tasks that supported the cover story. The purpose of the second session was to measure participants’ choice intentions, as well as the composition of their consideration sets. Participants were asked to assume that they were going to make a purchase within each of several categories, and to indicate which brand or store they would choose. They were then asked to list up to two additional brands or stores that they seriously considered prior to stating their choice intent. Participants recorded their choice intentions and other consideration set items on a blank sheet that did not list the names of any choice options. To evaluate H3 without bias, participants were randomly assigned to state a choice intention and indicate their consideration sets for only two out of the four categories that were included in the rehearsal task. H3 explored the potential of the rehearsal task to influence attitudes, and we were concerned that making a choice from a category might influence subsequently measured attitudes. Accordingly, by not having participants state choice intent and consideration set composition for two categories, the attitudes stated toward items within these categories would be uncontaminated by the choice task. Participants were asked to state choice intentions and other considered options for eight distractor categories in addition to the two choie categories of interest. The two relevant choice categories were randomly chosen from the four categories in the rehearsal task, as was the order of the choices.

After the choice intent and consideration set measures, participants’ attitudes towards all of the exemplars of each of the categories were measured. The order of exemplars on the attitude rating sheets was random, and accordingly Group A and Group B items were interspersed. The order of presentation of the categories was also randomized. Participants were asked to rank order their attitudes toward the exemplars belonging to each category.

Following completion of the choice intention and consideration set questionnaire, participants were given extra credit, debriefed, and excused.

TABLE 1

CONSIDERATION SET COMPOSITION AND CHOICE INCIDENCE AS A FUNCTION OF REHEARSAL CONDITION

TABLE 2

AVERAGE ATTITUDE TOWARD EXEMPLARS OF EACH CHOICE CATEGORY AS A FUNCTION OF REHEARSAL CONDITION

RESULTS

Table 1 presents consideration set and choice data. Because there were 106 participants, and choices were made from two categories, there were 212 possible choices. In addition to the choice from each category, participants were asked to list up to 2 additional options that were seriously considered but not chosen. Accordingly, in sum participants could have listed 424 options that were present in the consideration set but not chosen. In some cases participants did not make any choices from a category (e.g., "I change my own oil, so I wouldn’t go to any shop"), or failed to list 2 options that were considered but not chosen. As a result, in total there were 189 choices made, and 227 consideration set options listed (excluding the chosen option).

H1 was evaluated by comparing the number of options in the consideration set from the rehearsal group with the number of options in the consideration set from the control group. The first analysis considered options that were present in the consideration set, but were not chosen. H1 was supported as there were more items in participants’ consideration sets from the rehearsal group than from the control group, c2 (1) = 4.797, p < .03. Consistent with this result, when the data for the complete consideration set were analyzed (i.e., the complete consideration set includes choices and unchosen items that received consideration), more items from the rehearsal group were present than from the control group, c2 (1) = 8.654, p < .003.

H2 was evaluated by comparing the number of choices from the rehearsal group with the number of choices from the control group. The data show that options from the rehearsal group were more likely to be chosen than options from the control group, c2 (1) = 3.857, p < .05. Accordingly, rehearsing the association between a particular brand and the relevant superordinate choice category increased the likelihood that the brand would be chosen.

H3 would be supported by a finding of no difference in average attitude as a function of whether the category membership of exemplars was rehearsed vs. not rehearsed (see Table 2). To test H3 we computed within-subject t-tests comparing average attitude toward exemplars in the rehearsal vs. control group. Because we were concerned that stating a choice intention from a category might influence subsequently measured attitudes, these tests only included data for the two categories from which a participant did not make a choice. There was no evidence that the rehearsal task influenced attitudes as p values for each of the four product categories were all > .2.

DISCUSSION

The results of the experiment clearly show that manipulating the strength of association between brands and the choice category can have significant effects on subsequent memory-based consumer decision making. An alternative was much more likely to be present in consumers’ consideration sets if its category membership was rehearsed. Brands with experimentally strengthened associations to the choice category were also more likely to be chosen than control brands. This second finding is perhaps unsurprising given the first. In memory based choice, in which alternatives must be generated from memory prior to decision making, a brand cannot be chosen unless it is first identified as an alternative and included in the consideration set.

Consistent with earlier research, this study demonstrates that consideration effects in consumer decision making may operate independent of attitudes (cf. Nedungadi 1990). In this experiment, rehearsal of an alternative increased the probability that the alternative would be present in the consideration set and chosen, but not did influence attitudes.

The finding that rehearsal did not influence attitudes is important because it rules out a mere exposure account of the results. Research has shown that in some cases simply being exposed to stimuli increases individuals’ liking of the stimuli. If attitudes toward rehearsal set options were more favorable than attitudes toward control set options, one could argue that rehearsal led to attitude change, which in turn was reflected in decision making. The data, however, demonstrate that rehearsal did not influence liking of the choice options. Accordingly, it is implausible that mere exposure drove the findings.

This experiment, in conjunction with prior work (e.g., Kardes, et al. 1993; Nedungadi 1990; Posavac et al. 1997), demonstrates the importance of considering what is considered in consumer choice. Until recently, researchers in marketing and psychology generally explored decisions from constrained sets to the exclusion of contexts in which options must be generated from memory prior to choice. This research clearly demonstrates that manipulating a determinant of the likelihood that a brand will receive consideration (i.e., the strength of the brand-category association) may also influence the choice that is made when alternatives are not specified in context.

The managerial implications of this study are relatively straightforward. In line with prior research, we suggest that marketers should consider whether the brand being promoted belongs to a category relevant to decision contexts that contain vs. do not contain specified alternatives. In either case it is important to develop and maintain strong positive attitudes towards the target brand. If the brand to be promoted is typically chosen or not chosen in a memory-based context, an important second goal of promotion should be to develop and maintain a strong association between the brand and the relevant superordinate category. To the extent that the association is strong, the target brand is likely to be generated from memory upon consideration of the category, and accordingly will be eligible to be ultimately chosen.

Given the competitive advantage in memory-based choice engendered by a strong brand-category association, it is important to consider how this association may be strengthened. Generally, any intervention that results in consumers rehearsing the association between the target brand and the relevant category may increase the strength of the association. This experiment employed a simple rote rehearsal task. There are multiple methods of encouraging rehearsal available to practitioners. In a study of television ads Fazio, Powell, and Herr (1992) demonstrated that "mystery" television ads, in which the brand nd product category are not revealed until the end of the ad, contribute to strong category-brand associations. Fazio et al. suggest that the mystery format leads viewers to be cognitively ready to categorize the product when it finally appears, and this readiness leads to more elaborative rehearsal of the brand-category association. Other methods that may encourage category membership rehearsal include simply repeating the brand and the category in advertising, buying high advertising volume, creating a memorable tag line or jingle that mentions brand and category, and the dissemination of branded merchandise.

Caveats and Some Future Directions

The rehearsal task in this study was designed to increase the likelihood that target brands would be recalled upon consideration of the superordinate choice category. Farquhar and Herr (1993) termed the likelihood that a product category will evoke a particular brand the category dominance of the brand. These authors have made the extremely important, but somewhat under appreciated, distinction between category dominance and instance dominance. Instance dominance refers to the strength of the directional association from a brand to a particular product category. That is, given contact with a brand, how likely is it that a particular category will be evoked? Although both category and instance dominance are aspects of the representation of categories and brands in memory, category dominance is the more important determinant of a brand’s accession to the consideration set in memory-based choice. If a social influence attempt increases instance dominance instead of category dominance, the target brand would not likely enjoy the advantages of a strong category-brand association evident in this experiment. Future research on consideration effects in consumer choice should take account of this distinction.

We have suggested that strong category-brand associations may be particularly beneficial when alternatives are unspecified vs. specified in context. However, the benefits of a brand being strongly associated with the category may extend to stimulus-based choice contexts as well. In some instances individuals may form a choice intention for a product belonging to a stimulus-based category away from the point of purchase. For example, while constructing a shopping list a consumer may think of specific brands while noting the product categories he or she is in need of. A brand strongly associated with the category may pop into the consumer’s mind at this early stage of decision-making, and thereby be at an advantage when the consumer makes the stimulus-based choice. Future research is needed to delineate the circumstances in which strong brand-category associations are beneficial in stimulus-based choice.

REFERENCES

Farquhar, Peter H. and Paul M. Herr (1993), "The Dual Structure of Brand

Associations," in Brand Equity & Advertising, ed. David A. Aaker and Alexander L. Biel, Hillsdale, NJ: Lawrence Erlbaum Associates, 263-277.

Fazio, Russell H., Paul M. Herr and Martha C. Powell (1992), "On the Development and Strength of Category-Brand Associations in Memory: The Case of Mystery Ads," Journal of Consumer Psychology, 1, 1-13.

Kardes, Frank R., Gurumurthy Kalyanaram, Murali Chandrashekaren and Ronald J.

Dornoff (1993), "Brand Retrieval, Consideration Set Composition, Consumer Choice, and the Pioneering Advantage," Journal of Consumer Research, 20 (June), 62-75.

Lehman, Donald R. and Yigang Pan (1994), "Context Effects, New Brand Entry, and Consideration Sets," Journal of Marketing Research, 31 (August), 364-374.

Lynch, John G., Howard Marmorstein and Michael F. Weigold (1988), "Choices from Sets Including Remembered Brands: Use of Recalled Attributes and Prior Overall Evaluations," Journal of Consumer Research, 15 (September), 169-184.

Nedungadi, Prakash (1990), "Recall and Consumer Consideration Sets: Influencing Choice Without Altering Brand Evaluations," Journal of Consumer Research, 17 (December), 263-276.

Payne, John W., James R. Bettman and Eric J. Johnson (1993), The Adaptive Decision Maker. New York: Cambridge University Press.

Posavac, Steven S., David M. Sanbonmatsu and Russell H. Fazio (1997), "Considering the Best Choice: Effects of the Salience and Accessibility of Alternatives on Attitude-Decision Consistency," Journal of Personality and Social Psychology, 72 (February), 253-261.

Ratneshwar, S. and Allan D. Shocker (1991), "Substitution in Use and the Role of Usage Context in Product Category Structures," Journal of Marketing Research, 28 (August), 281-295.

Shapiro, Stewart, Deborah J. MacInnis and Susan E. Heckler (1997), "The Effects of Incidental Ad Exposure on the Formation of Consideration Sets," Journal of Consumer Research, 24 (June), 94-104.

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