The Impact of Initial Processing Goals on Memory-Based Brand Comparisons

ABSTRACT - When consumers compare two brands to form a preference, one brand serves as the starting point or subject of comparison and the other serves as the target or referent. Previous research has shown that the subject's unique attributes are more salient than the referent's unique attributes, and, consequently, consumers prefer the subject when both brands possess unique favorable attributes and they prefer the referent when both brands possess unique unfavorable attributes. The present experiment demonstrates that this direction-of-comparison effect is moderated by consumers' initial processing objectives. Stronger direction-of-comparison effects were found when respondents were instructed to memorize attributes than when they were told to form impressions of the products. The implications of the results for understanding memory-based judgment and choice are discussed.

Citation:

David M. Sanbonmatsu, Frank R. Kardes, and Bryan D. Gibson (1989) ,"The Impact of Initial Processing Goals on Memory-Based Brand Comparisons", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 429-432.

Advances in Consumer Research Volume 16, 1989    Pages 429-432

THE IMPACT OF INITIAL PROCESSING GOALS ON MEMORY-BASED BRAND COMPARISONS

David M. Sanbonmatsu, University of Utah

Frank R. Kardes, Massachusetts Institute of Technology

Bryan D. Gibson, University of Utah

ABSTRACT -

When consumers compare two brands to form a preference, one brand serves as the starting point or subject of comparison and the other serves as the target or referent. Previous research has shown that the subject's unique attributes are more salient than the referent's unique attributes, and, consequently, consumers prefer the subject when both brands possess unique favorable attributes and they prefer the referent when both brands possess unique unfavorable attributes. The present experiment demonstrates that this direction-of-comparison effect is moderated by consumers' initial processing objectives. Stronger direction-of-comparison effects were found when respondents were instructed to memorize attributes than when they were told to form impressions of the products. The implications of the results for understanding memory-based judgment and choice are discussed.

INTRODUCTION

Consumers commonly compare brand alternatives prior to making a purchase. Comparisons are performed to determine both similarities and differences between brands, and to determine which of the available brands is preferable. Common sense dictates that a comparison of two brands should be a straightforward matter of comparing all of the features of one brand with all of the features of the second brand. However, Tversky's (1977) feature-matching model of the comparison process predicts that certain asymmetries in judgment will result in a directional comparison.

Tversky suggests that in a comparison of two objects, one object often serves as the starting point or subject of comparison, and the second object serves as the target or referent. In Tversky's view, people performing a directional comparison begin with the features of the subject, and examine how the referent compares with regard to these features. Which object serves as the subject of comparison is critical, as people are particularly attuned to the subject's features. These features serve as a sort of checklist against which the features of the referent are reviewed. Features unique to the referent are often neglected in the comparison process, as people tend to compare only the features present in the subject. Features unique to the subject, however, tend to be highlighted by the comparison process and consequently carry the most weight in judgments about the two items.

A number of studies have demonstrated that the direction of comparison can lead to asymmetries in judgment consistent with Tversky's feature-matching model (Agostinelli, Sherman, Fazio, and Hearst 1986; Houston and Sherman forthcoming; Srull and Gaelick 1983; Tversky 1977). Of particular relevance to the present study is Houston and Sherman's examination of the impact of the direction of comparison on judgments of preference. The authors presented experimental participants with two different descriptions of objects of various categories (e.g., automobiles, apartments). The two competing objects were described by an equal number of favorable and unfavorable features. The descriptions differed, though, in that each object possessed features that were unique from the other object. In half of the conditions the objects possessed unique positive features and shared negative features, whereas in the remaining conditions the objects possessed unique negative features ag shared positive features.

Houston and Sherman (forthcoming) predicted that there would be a tendency to treat the most recently observed object as the subject, and the first observed and thus less accessible object as the referent. Three experiments supported this prediction. Further, this result is consistent with Biehal and Chakravarti's (1983) finding that when the most recently observed brand is acceptable, consumers choose this brand over other brands whose attributes are less accessible from memory. Thus, when respondents are exposed to two brands possessing equally favorable but unique features, respondents should prefer the subject of comparison over the referent. Mapping the features of the subject onto the referent should lead respondents to recognize the shared bad features and the subject's unique good features, and neglect the referent's unique good features.

Conversely, when two brands possess unique bad features, respondents should prefer the referent. Comparing two brands along the features of the subject should lead to a recognition of the common good features and the subject's unique bad features, and a neglect of the referent's bad features. Consistent with these predictions, Houston and Sherman found that the subject is preferred when the objects of comparison possess unique good features, and that the referent is preferred when the objects possess unique bad features. They also observed a pervasive tendency to treat the most recently observed object as the subject in a directional comparison.

The direction of comparison is presumed to affect judgments through a feature-matching process such as that proposed by Tversky (1977). Studies of memory-based judgment, though, indicate that consumers do not always rely on specific feature information in making a judgment (see Hastie and Park 1986; Kardes 1986; Lichtenstein and Srull 1985, 1987). When relevant judgments or impressions are formed during the initial processing of information, these impressions, rather than the specific information on which the impressions were based, tend to serve as the primary basis for later judgments. Thus, if consumers possess global impressions or evaluations of the relevant brands in memory, a memory-based comparison of the brands may be based on these global impressions rather than upon the specific feature knowledge of the brands.

We suggest that when a memory-based judgment is characterized by a comparison of brand impressions rather than brand attributes, the effects of the direction of comparison on judgments of similarity and preference observed in previous studies may be attenuated. When an impression is formed of the referent during the initial processing of information, both unique and shared features may be integrated into and accounted for in the overall brand impression or evaluation. The unique features of the referent may influence subsequent impression-based comparisons by affecting the nature of the initial global impression of the referent. Consequently, the tendency to overlook the referent's unique features in directional comparison may be attenuated.

The aim of the present study is to examine whether the initial processing of product information moderates the impact of the direction of comparison on memory-based preference judgments. We hypothesized that the formation of global brand impressions during the initial processing of brand information may attenuate the likelihood of the asymmetries in judgment that have been shown to occur as a function of the direction of comparison.

An experiment employing procedures and stimuli similar to those used by Houston and Sherman (forthcoming) was conducted. Respondents read about two brands from two different product categories (pens and automobiles). The brands were characterized by unique good features and shared bad features, or by unique bad features and shared good features. Respondents read about each brand and then were given a surprise comparison task. The initial processing of the products was manipulated such that half of the respondents read about the products under instructions to form an impression of the brands, whereas the remaining respondents read about the products under instructions to remember the attributes of the brands. It was predicted that the effect of the direction of comparison on preference observed by Houston and Sherman--the tendency to prefer the subject (i.e., the object presented second) more when both objects possessed unique good features as opposed to unique bad features--would be attenuated when respondents were prompted to form an impression.

METHOD

Stimulus Materials

Descriptions of two brands of cars and two brands of pens were developed. Each of the brands was described by five positive attributes (e.g., refillable, does not skip) and four negative attributes (e.g., not attractively styled, tip breaks with excessive pressure). Through pretesting, descriptions were created in which the five positive attributes describing one brand of a product category were equal in mean desirability to the five positive attributes describing the second brand. In addition, the four negative attributes describing one brand of a product category were equal in mean desirability to the four negative attributes describing the second brand. The automobile descriptions were developed by Houston and Sherman (forthcoming). Fictitious brand names were used for all four brands (HH Automobile, JJ Automobile, Dot Pen, Circle Pen).

Procedure

The experiment was described as "an investigation of people's perceptions of various consumer products." One hundred and thirteen subjects were given booklets containing the brand descriptions and were randomly assigned to impression or memory set conditions. In impression set conditions, respondents were instructed to "form an overall impression or evaluation of each of the products that are described. At the end of the session we will ask you a series of questions about your impressions." In memory set conditions, respondents were told to "remember the various attributes of each of the products that are described. At the end of the session, we will ask you to recall the various attributes of each of the products."

The descriptio s of the four brands were presented in a booklet. Each description was presented a separate page and consisted of a fictitious brand name followed by a list of the brand's features. Respondents read each of the four descriptions at a pace set by the experimenter. They were given a minute to read each description.

In unique good features conditions, respondents were exposed to descriptions in which the two brands of a product category shared bad features but had unique good features. In contrast, respondents in unique bad features conditions received descriptions of brands with shared good features and unique bad features.

The order of presentation of a brand description was counterbalanced, such that in half of the conditions one brand of a product category appeared prior to the other brand, whereas in the remaining conditions the order of presentation of the two brands was reversed. The product category appearing first in the booklet was also counterbalanced, such that in half of the conditions the first product described was an automobile, whereas in the remaining conditions the first product described was a pen. Finally, the descriptions were ordered such that the descriptions of the two brands of a product category never appeared consecutively.

Preference Measurement

After respondents read the product descriptions, the booklets were removed and a surprise preference measure was administered. Preference was assessed on a 12-point scale anchored by the endpoints "I strongly prefer the Brand HH Auto (Dot Brand Pen)" and "I strongly prefer the JJ Brand Auto (Circle Brand Pen)."

RESULTS

Because the pattern of results was similar for pens and automobiles, preference scores were collapsed across product category and a 2 X 2 (Processing goal X Valence of unique features) between-subjects analysis of variance was performed to test the prediction that respondents should prefer the subject of comparison over the referent when the unique features of the brands were positive rather than negative. As expected, a main effect for the valence of the unique features was obtained, as respondents preferred the subject more when the unique features were positive as opposed to negative. A main effect for processing set was not observed.

It was also predicted that the tendency to prefer the subject of comparison more when the features of the brands were positive as opposed to negative would be more pronounced in memory set than in impression set conditions. The expected interaction was marginally significant. Planned comparisons revealed that participants under memory set conditions preferred the subject more in unique positive than in unique negative features conditions. A similar but weaker pattern was observed in impression set conditions. Thus, support was found for the hypothesis that the direction of comparison will have a greater impact on preferences in memory set than in impression set conditions.

DISCUSSION

When two brands possess shared bad features and unique good features, consumers prefer the subject of comparison (i.e., the brand presented second). In contrast, when two brands possess shared good features and unique bad features, consumers prefer the referent (i.e., the brand presented first). This direction-of-comparison effect occurs even when the mean desirability of the unique features are equivalent for the two brands. The present study demonstrated that this effect is very pervasive: it occurs across product categories and it occurs in memory set and in impression set conditions. However, direction of comparison had a greater impact on preference in memory set than in impression set conditions, as predicted.

According to Tversky (1977), when people compare objects the features of the subject are mapped onto the features of the referent. As a consequence, people are especially likely to detect differences between two objects when the subject possesses features that are not shared by the referent. On the other hand, the referent's unique features are likely to be overlooked because the subject--not the referent-determines what features are examined.

The present study, however, demonstrates that the referent's unique features can influence preference under some circumstances. For example, when consumers are prompted to use the referent's unique features to form an overall impression of the referent, consumers are less likely to overlook these features. We argued that this is because consumers may use shared and unique features to form an overall impression of a brand and then a preference judgment might be made by comparing overall impressions between two brands. An alternative explanation, however, is that an impression set may enhance recall for all features for both brands (Hamilton, Katz, and Leirer 1980; Wyer and Srull 1986) and an attribute-based comparison process may be used to form a preference judgment. With either process, however, an impression set increases the likelihood of considering the referent's unique features and the magnitude of the direction-of-comparison effect is reduced.

Because the unique features of the subject have a strong impact on preference, it is important to identify the factors that determine which brand will be the subject and which brand will be the referent. The present study and the Houston and Sherman (forthcoming) studies show that order of presentation influences the direction of comparison. When two brands are presented sequentially, the second brand is the most recently considered brand. Because recently activated information is readily accessible from memory, the features of the second brand are more accessible than the features of the first brand, and, consequently, the second brand is used as the starting point or subject of comparison.

However, recency of activation is not the only determinant of information accessibility. Frequency of activation is another important determinant (Higgins and King 1981, Srull and Wyer 1986). If the most accessible brand in a product category is used as the subject of comparison, it could be predicted that frequent exposure to attribute information for one brand may induce consumers to use that brand as the subject. If so, the attributes of this brand should be salient and the attributes of competing brands should be overlooked to some extent.

It is easy to think of naturally occurring situations in which this process may take place. For example, consumers are often exposed to the features of a pioneer brand over a period of time. By the time competing brands enter the market, the features of the pioneer brand should be readily accessible from memory. If so, the pioneer brand should be used as the subject of comparison. The unique features of later entrants may be overlooked by consumers, and, consequently, competitors may experience difficulty in differentiating their brands from the pioneer brand. Hence, the direction-of- comparison effect may provide one explanation for the pioneer brand advantage (see Urban et al. 1986). We are currently pursuing this line of investigation.

REFERENCES

Agostinelli, Gina, Steven J. Sherman, Russell H. Fazio, and Eliot S. Hearst (1986), "Detecting and Identifying Change: Additions Versus Deletions," Journal of Experimental Psychology: Human Perception and Performance, 12 (April), 445-454.

Biehal, Gabriel and Dipanker Chakravarti (1983), "Information Accessibility as a Moderator of Consumer Choice," Journal of Consumer Research, 10 (June), 1-14.

Hamilton, David L., Lawrence B. Katz, and Von O. Leirer (1980), "Cognitive Representation of Personality Impressions: Organizational Processes in First Impression Formation," Journal of Personality and Social Psychology, 39 (December), 1050-1063.

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

Higgins, E. Tory and Gillian King (1981), "Accessibility of Social Constructs: Information-Processing Consequences of Individual and Contextual Variability," in Personality, Cognition, and Social Interaction, eds. Nancy Cantor and John F. Kihlstrom, Hillsdale, NJ: Lawrence Erlbaum Associates, 69-121.

Houston, David A. and Steven J. Sherman (forthcoming), 'The Influence of Unique Features and Direction of Comparison on Preferences," Journal of Experimental Social Psychology, in press.

Kardes, Frank R. (1986), "Effects of Initial Product Judgments on Subsequent Memory-Based Judgments," Journal of Consumer Research, 13 (June), 1-11.

Lichtenstein, Meryl and Thomas K. Srull (1985), "Conceptual and Methodological Issues in Examining the Relationship Between Consumer Memory and Judgment," in Psychological Processes and Advertising Effects: Theory, Research, and Application, eds. Linda F. Alwitt and Andrew A. Mitchell, Hillsdale, NJ: Lawrence Erlbaum Associates, 113-128.

Lichtenstein, Meryl and Thomas K. Srull (1987), "Processing Objectives as a Determinant of the Relationship Between Recall and Judgment," Journal of Experimental Social Psychology, 23 (March), 93-118.

Srull, Thomas K. and Lisa Gaelick (1983), "General Principles and Individual Differences in the Self as a Habitual Reference Point: An Examination of Self-Other Judgments of Similarity," Social Cognition, 2, 108-121.

Tversky, Amos (1977), "Features of Similarity," Psychological Review, 84 (July), 327-352.

Urban, Glen L., Theresa Carter, Steven Gaskin, and Zofia Mucha (1986), "Market Share Rewards to Pioneering Brands: An Empirical Analysis and Strategic Implications," Management Science, 32 (June), 645-659.

Wyer, Robert S. and Thomas K. Srull (1986), "Human Cognition in Its Social Context," Psychological Review, 93 (July), 322-35

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