A Model of Brand Evaluation Formation With Memory Based Context Effects

Peter A. Dacin, Queen=s University
Andrew A. Mitchell, University of Toronto
[ to cite ]:
Peter A. Dacin and Andrew A. Mitchell (2002) ,"A Model of Brand Evaluation Formation With Memory Based Context Effects", in NA - Advances in Consumer Research Volume 29, eds. Susan M. Broniarczyk and Kent Nakamoto, Valdosta, GA : Association for Consumer Research, Pages: 458-459.

Advances in Consumer Research Volume 29, 2002     Pages 458-459

A MODEL OF BRAND EVALUATION FORMATION WITH MEMORY BASED CONTEXT EFFECTS

Peter A. Dacin, Queen=s University

Andrew A. Mitchell, University of Toronto

A model of brand evaluation formation is presented which builds on the literature on the priming of exemplars on unambiguous targets (e.g. Herr, 1986, 1989; Schwartz and Bless, 1992), exemplar based judgment models (e.g. Kahneman and Miller, 1986) and attitude theory (e.g. Fishbein and Ajzen, 1975). The model contains five stages. At the first stage consumers use the presented information about the brand to form a cognitive representation of the brand and then categorize it. Next, the consumer recruits highly accessible alternative brands from that category from memory. Third, standards of comparisons are generated for the attribute values of the recruited brands. At the fourth stage, the consumer compares the attribute levels of the target brand to the standards of comparison and determines the evaluation of the attribute level. Finally, the consumer uses the attribute evaluations to form an overall evaluation of the target brand.

Empirically, this means that if the attribute values of the recruited brands are poorer than the attribute values of the brand which is being evaluated, the attitude evaluations of the latter brand are positive, as will be the brand attitude. Alternatively, if the attribute values of the recruited brands are better than the brand to be evaluated, the attribute evaluations will be negative, as will the brand attitude.

There are six critical aspects of the model which need to be tested. The first is that a number of highly accessible brands in memory from the same category will be recruited to form a standard of comparison which will create contrast effects on the evaluation of an unambiguous target brand from the same product category. Secondly, the recruitment process is an activation as opposed to an inhibitory process. Third, the generation of alternative brands in memory with an unambiguous target brand affects the evaluation of the attribute levels of that brand. Fourth, the sum of the evaluations of the attribute levels predicts the resulting evaluation of the target brand. Fifth, accessible brands are recruited to form a standard of comparison even if they have not been primed. Finally, the recruitment process, the formation of standards of comparison, and the formation of attribute levels and their evaluations require few conscious cognitive resources. The first two aspects will be tested in Experiment 1, the second two in Experiment 2, the fifth in Experiment 3 and the sixth in Experiment 4.

Since the same brands will be typically recruited for the evaluation of a brand, in order to test the model it is necessary to make brands accssible that would not ordinarily be accessible, or to create new brands and make them accessible. Therefore, in the experiments, two types of manipulations will be used. First, existing brands will be primed. Second, subjects will learn about new brands. We refer to the former as primed brands and the latter as context brands.

The first experiment is a 3 x 2 x 2 between subject design. The manipulations included three levels of Context Brand (Highly Evaluated Context Brand, Moderately Evaluated Context Brand and No Context Brand-Control), two levels of Label (Consistent Label and Inconsistent Label) and two unambiguous Target Brand descriptions.

The results of the first experiments are consistent with the predictions of the model. The context brands create contrast effects on the brand evaluations in both the consistent and inconsistent label conditions. Since the inconsistent label is for a poor subcategory of the category product, it creates more positive attitudes toward the target brand. This indicates that the inconsistent label initially causes brands from that category before the subjects realize that the label is not consistent with the description of the target brand. This indicates that multiple brands were recruited and the recruitment process is an activation not an inhibitory process.

The second study is a 2 x 3 between subject. The manipulations in this design include three levels of Prime (Highly Evaluated Prime, Moderately Evaluated Prime and No Prime), and two levels of Ambiguity of the Target Brand (Ambiguous Stimulus and Unambiguous Stimulus).

The results indicate that the recruitment of the alternative brands with an unambiguous target produced contrast effects on the attribute evaluation and the brand attitudes, as predicted by the model, while the recruitment of alternative brands with an ambiguous target produced assimilation effects on both the perceived attribute levels and attribute evaluations. In addition, the results indicate that the sum of the attribute evaluation predicts the overall evaluation of the target brand.

The third experiment is a 2 x 2 between subjects design with two levels of Context Brand (High Acceleration/Low Fuel Economy and Low Acceleration/High Fuel Economy), and two levels of Timing of Measures (Immediate and Two-Day Delay). The target brand is unambiguous.

The results indicate that the context brands produced contrast effects on the attribute evaluation in the immediate condition and in the two-day delay condition. No differences were found between the immediate and the two-day delay conditions. This suggests that the same brands were recruited to form a standard of comparison in both the immediate and delay conditions and that the recruited brands do not have to be primed to produce contrast effects.

The fourth experiment is also a 2 x 2 between subjects design with two levels of Context Brand (High Acceleration/Low Fuel Economy, Low Acceleration/High Fuel Economy), and two levels of Memory Task (Memory Task and No Memory Task). Care is taken in designing the memory task, so that subjects can read the information about the target brand, but cannot do further processing of it.

Again, the context brands produced contrast effects on the attribute evaluations in both the Memory Load and No Memory Load conditions. There were no significant differences in the attribute evaluations between Memory Load and No Memory Load conditions. This indicates that little cognitive effort is required to generate alternatives, form standards of comparisons, attribute evaluations and attitudes.

In summary, the results of the four experiments provide strong support for the model.

REFERENCES

Fishbein, Martin and Icek Ajzen (1975), Belief, Attitude, Intention and Behavior, Reading, MA: AddisonBWesley.

Herr, Paul M.(1989),"Priming Price: Prior Knowledge and Context Effects," Journal of Consumer Research, 16 (June), 67-75.

Herr, Paul M. (1986), "Consequences of Priming: Judgment and Behavior," Journal of Personality and Social Psychology, 51 (December), 1106-1115.

Kahneman, Daniel and Dale T. Miller (1986), "Norm Theory: Comparing Reality to Its Alternatives," Psychological Review, 93 (April), 136-153.

Schwartz, N. and H. Bless (1992) "Constructing Reality and its Alternatives: Assimilation and Contrast Effects in Social Judgment", in L.L. Martin and A. Tessor (eds.) The Construction of Social Judgments, Hillsdale, N.J.: Erlbaum.

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