The Influence of Contextual Priming on Advertising Effects

Youjae Yi, University of Michigan
ABSTRACT - This study investigates a particular way in which contextual priming influences advertising effects. It is proposed that prior exposure to contextual factors can prime or activate certain product attributes in consumers' knowledge structure and subsequently increase the likelihood that they interpret ambiguous product information in terms of these activated attributes, thereby affecting the overall impact of the ad. Two experiments are conducted to test the main hypothesis and eliminate potentially confounding effects of experimental tasks. The results demonstrate that the specific attributes relevant to evaluating the advertised brand vary in their accessibility as a function of the ad context, and that these variations influence brand attitudes and purchase intentions. Step-down analyses show further that the effects of priming product attributes operate mainly through brand attitudes. Theoretical and practical implications of these findings are also discussed.
[ to cite ]:
Youjae Yi (1991) ,"The Influence of Contextual Priming on Advertising Effects", in NA - Advances in Consumer Research Volume 18, eds. Rebecca H. Holman and Michael R. Solomon, Provo, UT : Association for Consumer Research, Pages: 417-425.

Advances in Consumer Research Volume 18, 1991      Pages 417-425

THE INFLUENCE OF CONTEXTUAL PRIMING ON ADVERTISING EFFECTS

Youjae Yi, University of Michigan

ABSTRACT -

This study investigates a particular way in which contextual priming influences advertising effects. It is proposed that prior exposure to contextual factors can prime or activate certain product attributes in consumers' knowledge structure and subsequently increase the likelihood that they interpret ambiguous product information in terms of these activated attributes, thereby affecting the overall impact of the ad. Two experiments are conducted to test the main hypothesis and eliminate potentially confounding effects of experimental tasks. The results demonstrate that the specific attributes relevant to evaluating the advertised brand vary in their accessibility as a function of the ad context, and that these variations influence brand attitudes and purchase intentions. Step-down analyses show further that the effects of priming product attributes operate mainly through brand attitudes. Theoretical and practical implications of these findings are also discussed.

INTRODUCTION

Product information in advertisements can often have multiple implications for the evaluation of the advertised product. For example, when an ad emphasizes that a car is large, one might infer either that the car will be comfortable or that the car will yield low miles per gallon (MPG). In such a case, interpretations of the large size as high comfort would induce favorable brand evaluations, whereas interpretations as low gas mileage would yield unfavorable brand evaluations. A question then arises: What determines the particular interpretation given to the product information that has several possible meanings?

According to the research on priming, interpretations of ambiguous information can be very sensitive to the surrounding context (e.g., Herr 1989; Meyers-Levy 1989). Researchers have found that an ambiguous stimulus is often interpreted in terms of the concepts that are primed by contextual factors (Wyer and Srull 1981). These findings suggest that contextual factors might be an important determinant of how ambiguous information in the ad is perceived.

Thus, the present study proposes that contextual factors can affect consumers' processing of ambiguous product information in the ad by priming certain attributes. Specifically, contextual materials (e.g., a magazine article on oil prices) may activate particular product attributes (e.g., gas mileage) and guide consumers' interpretations of product information (e.g., car size). These interpretations may result in the formation or change of beliefs about the advertised brand, thereby affecting consumers' evaluations of the advertised brand.

CONTEXTUAL PRIMING EFFECTS

Researchers using the priming paradigm have shown that people's interpretation of information often depends on the particular knowledge structures (e.g., concepts and schemas) that are currently active (Higgins and King 1981; Wyer and Srull 1981). For example, the fact that someone gave a friend an answer during an exam could be interpreted as either "dishonest" or "kind." The type of interpretation given seems to depend on which of the related concepts (dishonest or kind) is easily accessible at the time when information is processed (Srull and Wyer 1980). Accessible concepts serve to direct attention to certain aspects of information and are likely to guide the interpretation of information (Yi 1990a). Priming effects have been demonstrated even when people are unaware of the activated concepts (Higgins, Bargh, and Lombardi 1985). These findings suggest that highly accessible attributes are likely to be used in interpreting an ambiguous description of a product.

Then, what makes certain attributes highly accessible? Many researchers in cognitive and social psychology have found that the accessibility (likelihood of retrieval from memory and subsequent use) of a certain concept is enhanced by prior exposure to the concept (Higgins and King 1981; Wyer and Srull 1986). A concept's temporary accessibility is directly related to its recency of activation; the more recently a concept is activated, the greater its accessibility (Higgins and King 1981). Although a variety of factors can provide exposure to product attributes, of particular interest to this study is the context for the ad.

Many advertisements occur with other materials such as articles in the magazine and other competing advertisements (Soldow and Principe 1981), and prominent aspects of the ad environment can activate certain sets of product attributes (Gardner 1983). When the ad context provides people with exposure to a certain attribute, such as occurs when they read a magazine article emphasizing the attribute, this attribute may become highly accessible. Subsequently, that attribute is likely to be used in processing the ad information and evaluating the advertised brand. That is, contextual factors may make certain attributes salient to ad recipients and guide their perceptions of product information in the ad (Shavitt and Fazio 1990).

The evaluation of the advertised brand would therefore depend upon which attribute is activated by the contextual factors preceding the ad. When the ad context primes an attribute (e.g., ease of handling) that has positive implications for evaluation of the target brand, overall product evaluations will be enhanced. In contrast, when the context primes an attribute (e.g., durability) whose evaluative implication is negative, overall product evaluations will be lowered. This suggests that the same ad can have different effects, depending upon the attribute activated by the context preceding the ad. Thus, it is hypothesized that how product information in an ad is perceived depends on what related attribute is primed previously by contextual factors.

One purpose of the present study is to investigate the priming effects of contextual factors for a print advertisement, especially when the ad contains ambiguous product information. The results should be informative of a way in which the ad context can influence ad perception. Another purpose of this research is to look deeply into the nature and process of contextual priming effects. In this regard, we analyze attribute level data elicited by subjects to assess the accessibility of primed attributes and examine the theoretical relation among dependent variables with step-down analyses. We have also employed the more powerful analysis of MANOVA designs via structural equation models (Bagozzi and Yi 1989). Finally, an attempt is made to replicate the results so that the findings can be generalized. By conducting a second experiment that overcomes certain methodological limitations of the first experiment, we attempt to eliminate alternative explanations and test the generalizability of the findings.

EXPERIMENT 1

The priming hypothesis was tested in an experiment in which subjects were exposed to an ad and asked to indicate their reactions to the advertised product. The contextual materials surrounding the target ad were manipulated to investigate the impact of the ad context. All the subjects saw the same target ad, except for the context.

Method

Subjects and Advertisement. Forty student subjects participated in the first experiment. Personal computers were selected as a focal product, because this product category should be highly relevant to the subjects. After a pretest was conducted with ten students, the availability of various features was chosen as the key message in the target ad. Accordingly, ease of use was chosen as the attribute to be primed in the negative interpretation condition, whereas versatility was chosen as the salient attribute in the positive interpretation condition (see also Yi 1990d).

The target ad focused on numerous features of a new personal computer (PC-3000). The ad featured a headline in bold face, "Get the PC-3000, satisfy your lust for power and performance." The ad contained three paragraphs of text emphasizing that the PC-3000 personal computer has numerous features. The ad also contained a picture of one person standing by the personal computer.

Procedure. After being seated in the research room, subjects were told that the study concerned consumers' evaluation of print advertisements for new personal computers in pre-production form. After the general instructions, they were given three booklets and told to complete the booklets in the order presented.

In the first booklet, subjects were asked for general background information such as their knowledge and familiarity with personal computers. The second booklet contained two advertisements: (1) an ad designed to prime certain product attributes and (2) an ad for the target brand (PC-3000). The first ad in this booklet represented the priming manipulation. The purpose of this ad was to prime a certain product attribute that could induce an either positive or negative interpretation of the advertised attribute in the target ad ("numerous features" of PC3000). To do this, two different ads were created that emphasized one of the two attributes (versatility and ease of use). These ads will be called hereafter "prime ads." Each of the prime ads featured a personal computer (which is different from the target brand) that can activate one of the attributes relevant to evaluating the target brand. An ad was used as a prime, because it fit well with the cover story of the experiment (cf. Yi 1990b).

Half of the subjects saw a prime ad for the personal computer with a brand name in bold face, "Versa-Com." Centered at the top of the prime ad was the headline, "I didn't know it could do that," emphasizing the fact that the Versa-Com computer can perform many fictions. This claim was bolstered by including a picture illustrating many different analyses and applications of the computer. The other half saw a prime ad emphasizing the ease of use for the computer with its brand name in bold face, "EZ-Com." This prime ad had the headline, "Our frills require no skills," at the top and contained a picture of a child working at the computer terminal. Each of the two prime ads had the same format: a headline, two paragraphs of copy text, and a picture. Both ads were approximately equal in length.

The objective of the priming manipulation was to enhance the likelihood that, later when reading the target ad mentioning that the PC-3000 computer had numerous features, subjects who had earlier seen the Versa-Com ad would encode the provided information in terms of versatility, whereas those who had earlier read the EZ-Com ad would perceive the product information in terms of ease of use. After completion of the priming manipulation, subjects saw the target ad for PC-3000. All subjects saw the same target ad, although they had previously read a different prime ad (either Versa-Com or EZ-Com).

In the third booklet, subjects were asked to generate salient attributes of a personal computer that would come to mind if they considered purchasing a personal computer. They listed the characteristics of a personal computer that they would consider in an open-ended format. Subjects were then asked to turn to the next page, where they were asked for brand evaluations such as brand attitude and purchase intentions. It was also checked whether there had been any demand artifacts. After completing the last questionnaire, subjects were asked to write down their thoughts concerning the purpose of the study. Results showed that no subjects guessed the true purpose of the study, indicating that few demand artifacts had operated.

Dependent Variables. Attitude toward the target brand (Ab) was assessed by three 7-point scales anchored by the phrases "good-bad," "like-dislike," and "favorable-unfavorable." Purchase intention toward the brand (PI), assuming a product category need, was also measured by asking a question, "What are your chances of buying PC-3000 the next time that you need to purchase a personal computer?" Given a single exposure to the ad, the conditional purchase intention, rather than the usual purchase intention, measure was considered appropriate. Subjects responded on three seven-point scales: "likely-unlikely," "possible-impossible," and "probable-improbable." The alpha coefficients for Ab and PI were .90 and .92, respectively, indicating a high degree of internal consistency.

Results

Contextual Priming Effects. A one-way multivariate analysis of variance (MANOVA) was run via SPSSX on the set of two dependent variables (Ab and PI), each of which was operationalized by the sum of three items. The Box's M test indicated that the homogeneity assumption was valid: Box's M = 1.93, x2 (3) = 1.82, p > .60. MANOVA results showed that contextual priming had a significant effect on these measures of brand evaluations (F (2, 37) = 7.92, p < .01).

For an understanding of priming effects on individual variables, separate ANOVAs were subsequently run on each dependent variable. Results indicated that priming had significant effects on both Ab and PI (F (1, 38) = 12.59, p < .01; F (1, 38) = 10.74, p < .01, respectively). As expected, Ab was higher when the versatility attribute was primed, compared with the case when ease of use was primed (4.98 vs. 3.97). Also, BI was higher in the versatility condition (4.57) than in the ease-of-use condition (3.33).

We have thus far examined results of MANOVA and univariate ANOVAs for priming effects. There are, however, several limitations to this common use of MANOVA-ANOVA analyses (Bray and Maxwell 1985). First, the probability statements from separate ANOVAs are not meaningful, when the dependent variables are interrelated. Previous research shows that the dependent variables (Ab and BI) of this study are related to each other (e.g., Mackenzie, Lutz, and Belch 1986). Second, MANOVA or ANOVAs are not very useful for understanding the nature and process of the experimental effects on interrelated variables as in this study. Variation in a particular variable may be either due to a direct influence of the priming manipulation or due to the dependence of that particular variable on other variable.

A useful approach to such problems would be a step-down analysis (Roy 1958). The step-down analysis provides an examination of sequential relations among the set of dependent variables. By examining dependent variables in a predetermined way, one can assess the unique contribution of each variable to the between-group difference, as the variable is added to the dependent variable set. Unlike univariate ANOVA tests, the probability values associated with step-down analysis are independent. It can provide useful information by testing whether variation in a certain dependent variable is due to the direct effect of the experimental manipulation or due to the relationships of that dependent variable with other dependent variables.

Previous research suggests that the dependent variables are likely to be in a causal order of Ab to PI (e.g., Mackenzie, Lutz, and Belch 1986; Yi 1990b). Step-down analyses were thus conducted with a causal path from Ab to PI. Step-down analyses were conducted beginning with the last ordered variable (Ab), examining the step-down F values. The first step-down F value was the same as the univariate F value from ANOVA on Ab (F (1, 38) = 12.59, p < .01). But the next step tested the effect of contextual priming on PL with the effect of Ab covaried out. The results showed that the effect of priming was statistically insignificant (F (1, 37) = 2.68, p > .10), whereas the effect of Ab as a covariate was significant (F (1, 37) = 19.08, p < .01).

Thus, when the causal relation between the dependent variables was taken into account by step-down analyses, one of the effects that had been significant in univariate ANOVAs became insignificant. Specifically, the priming effect on PI became insignificant in the step-down analyses. These results suggest that the variation in PI was due to the dependence of PI on Ab, rather than due to the direct influence of priming itself.

Attribute Accessibility. We have thus far examined the overall brand evaluations, and implicitly assumed that if priming effects on judgements were observed, the attribute must have been activated. However, since priming effects depend ultimately on the enhanced accessibility of a primed attribute, it is important to assess whether different attributes were indeed accessible to consumers across the two priming conditions. In this regard, the attribute data from the elicitation task were analyzed to gain better insights into the processes underlying the priming effects.

Two measures were constructed from the free elicitation data to operationalize the accessibility of attributes: (1) frequency of mention and (2) order of mention (Yi 1990b). The frequency of mention measures were based on the assumption that accessible attributes would be more frequently mentioned by subjects (Ryan and Holbrook 1982). The order of mention measure was used, because it is likely that a cognitively accessible concept comes first to mind (Jaccard, Brinberg, and Ackerman 1986; Wyer and Srull 1981). This measure is also consistent with the availability heuristic (Tversky and Kahneman 1973).

The frequency of mention measure was first examined. For the attribute of versatility, the priming manipulation had significant effects on the frequency of mention; 55% of the subjects mentioned versatility in the versatility condition, whereas 25% mentioned versatility in the ease-of-use condition (p < .03). On the other hand, ease of use was more frequently mentioned in the ease-of-use condition than in the versatility condition (55% vs. 30%; p < .06). The order of mention measure was also compared across groups. For the attribute of versatility, the mean order of mention was 6.2 in the versatility condition, compared with 7.1 for the ease of use condition (Mann-Whitney U = 142.5; p <.05). The ease of use was also elicited earlier in the ease-of-use condition than in the versatility condition (6.0 vs. 7.7; Mann-Whitney U = 140.5; p < .05). The results indicate that the priming manipulation indeed affected the relative accessibility of product attributes.

Overall, the results of Experiment 1 supported the main hypothesis. Evaluations of the target brand were influenced by the ad context priming different product attributes. Step-down analyses also showed that contextual priming affected Ab directly, but affected PI indirectly (through Ab). Further, attribute level data indicated that primed attributes were indeed highly accessible to consumers as expected.

Limitations. We can note several potential limitations of the first experiment. One problem concerns the requirement after reading the target ad that subjects generate salient attributes of a personal computer that would come to mind if they consider purchasing a personal computer. This task might have been intrusive and have biased the subjects' processing of the product information in such a way as to support the predictions of the priming hypothesis. That is, the task might have drawn the subjects' attention to these attributes, emphasizing them in a way unlike what would naturally occur. It is thus possible that the effects on Ab and PI might have been just the result of the attribute eliciting task, rather than the priming manipulation. To the extent that such a possibility exists, the interpretation of the results is ambiguous.

Second, the dependent variables (Ab and PI) in this study are inherently unobservable and they might have to be treated as latent constructs in analyses. In this regard, it seems instructive to test the hypothesis via more powerful analyses with structural equation models (Bagozzi and Yi 1989). However, the small sample size in this experiment does not permit us to do this analysis.

EXPERIMENT 2

Method

A second experiment was conducted to correct for the aforementioned limitations of the first experiment. The second experiment differed from the first experiment in two ways. First, the experiment eliminated the attribute elicitation task, which could be intrusive so as to distort the subjects' processing of ad information. Second, the sample size was large enough for structural equation analysis; the subjects were 120 students recruited from several business courses. The stimuli and procedures were identical to those of the first experiment in all other aspects.

Results

A one-way MANOVA was run first via SPSSX on the set of two dependent variables (Ab and PI), each of which was operationalized by the sum of three items. The Box's M test indicated that the homogeneity assumption was valid: Box's M = 4.34, x2 (3) = 4.26, p > .23. MANOVA results showed that attribute priming had a significant effect on these measures of advertising effectiveness (F (2, 117) = 4.91, p < .01). For an understanding of priming effects on individual variables, separate ANOVAs were also run on each dependent variable. Contextual priming had significant effects on both Ab and PI (F (1, 118) = 7.93, p < .01; F (1, 118) = 4.26, p < .04, respectively). An examination of the cell means reveals that the effects were in the expected direction. Both Ab and PI were higher when the versatility attribute was primed, compared with the case when the ease of use was primed (4.56 vs 3.98; 3.82 vs 3.32, respectively).

Note that F values and mean differences for the priming effects are lower in Experiment 2 than in Experiment 1 (e.g., MANOVA F = 7.92 vs 4.91). This is consistent with our prediction that the task of attribute elicitation might be intrusive so as to increase priming effects. However, the priming effects were still significant in Experiment 2 which did not involve this task. These results suggest that the attribute elicitation task cannot fully explain the mean differences between the two groups, and that contextual priming itself has significant effects on brand evaluations.

Step-down analysis was then conducted via structural equation models (Bagozzi and Yi 1989). This analysis enables one to employ latent constructs indicated by several items as the dependent variables. The augmented moment matrix, rather than the correlation or covariance matrix, was used as the input data for analyses with LISREL. Figure 1 illustrates the specifications of the structural equation models for the step-down analyses.

The initial stage was a MANOVA test performed on the dependent variables, which were operationalized as latent constructs underlying the observed variables (see Figure 1A). For example, Ab was used as a factor underlying the three measured variables (Ab1, Ab2, and Ab3). Notice that the two experimental groups were represented by a 0, 1 dummy variable, which was expressed as an exogenous latent variable (41). Note also that a pseudo-variable (i.e., "one") was shown as another exogenous variable (E,2) to capture the means or locations of dependent variables. Because the dummy variable was 0 for one group and 1 for the other group, the paths from the dummy variable to dependent variables corresponded to the differences in the means across the two groups. Specifically, Y1 and Y2 were the mean differences between the two groups in Ab and PI, respectively. The global significance of the mean differences were tested with the chi-square difference tests of the zero restrictions for these parameters.

TABLE 1

FINDINGS FOR STEP-DOWN ANALYSIS VIA STRUCTURAL EQUATION MODELS IN EXPERIMENT 2

The top portion of Table 1 reports the findings for the initial stage of the step-down analysis: the omnibus test with all variables included but no causal relation implied between them. The full model specified in Figure 1A, which allows for the differences in means, gave the following results: x2 (16) = 32.49, p = .009. The mean differences parameters (i.e., Y1 and t2) were 0.55 (t = 2.78) and 0.46 (t = 2.00). The restricted model with the zero constraints for the mean difference parameters gave the following results: x2 (18) = 41.63, p = .001. The chi-square difference was 9.14 with 2 degrees of freedom, which was significant at the .02 level. These findings suggested that the means of dependent variables were different across groups.

The next two steps consisted of testing the mean differences while controlling for the theoretical relation between dependent variables (see Figure 1B). In Step Two, the mean difference in PI was tested after considering the causal order between Ab and PI. In other words, t2 in Step Two could be interpreted as the mean difference in PI due to the priming manipulation when the effect of Ab had been controlled for. The chi-square difference test indicated that the mean difference in PI was not significant: X2d (1) = 1.49, p > .10. In the final step, the mean difference in Ab was tested for statistical significance. The chi-square difference regarding this mean difference was significant: X2d (1) = 7.65, p < .01.

When the causal relation among the dependent variables was taken into account, one of the effects that had been significant in univariate ANOVAs became insignificant. Specifically, the effect of attribute priming on PI became insignificant in the step-down analyses. These results suggest again that the variation in PI was due to the dependence of PI on Ab, rather than due to the direct influence of priming itself. In sum, these results replicate the findings of Experiment 1; the contextual priming of product attributes affected Ab directly, but influenced PI indirectly through Ab.

DISCUSSION

It is found that priming a particular attribute increases the likelihood that this attribute will be used to interpret product information in an ad, and thus influences the evaluation of the advertised brand. The activation of the attribute guided a person to select among possible interpretations of an ambiguous description (i.e., the number of features) of the target brand. When subjects were reading the prime ad on either of the two attributes (versatility or ease of use), this attribute should have been activated and become highly accessible. As a consequence, subjects should have had the "top of mind" awareness of the attribute when they subsequently read the target ad. That attribute was therefore highly likely to be used in processing product information in the ad. That is, the interpretation given depended on which attribute was most accessible when ambiguous information was received.

FIGURE 1

STRUCTURAL EQUATION MODELS FOR STEP-DOWN ANALYSIS

These results support research in social cognition showing that construct accessibility can increase temporarily from recent activation and affect people's judgment of an object (Wyer and Srull 1981). According to Wyer and Srull's "storage bin" model, a recently activated concept is placed on the top of a layered bin, and the construct at the top is most likely to be used in interpreting new incoming information. The elicitation data in Experiment 1 indicate that the attributes primed by the preceding context were indeed accessible for use in interpreting ad information. The results are also in line with the research on framing effects in consumer decision making (e.g., Bettman and Sujan 1987); priming different decision criteria (i.e., attributes) influences how a product is evaluated.

This research extends existing studies in several aspects. First, this study incorporates research on priming in investigating the effects of contextual materials preceding the ad (Herr 1989; Meyers-Levy 1989). This study has found that contextual factors may influence judgments of the advertised product by altering the way how information is perceived. The same product features in an ad can be evaluated in different ways, depending on the adjacent materials. This finding is consistent with the Gestalt psychology, decision making, and perception stressing that the context in which a stimulus appears affects the interpretation of that stimulus (e.g., Helson 1964; Payne 1982).

Second, this study links research on information accessibility and ad context effects within a single framework. On the one hand, researchers have found that information accessibility affects brand choice and attitudinal judgments (e.g., Biehal and Chakravarti 1983; Kisielius and Sternthal 1986). On the other hand, many studies have shown that ad contexts affect advertising effectiveness (e.g., Chook 1985; Soldow and Principe 1981). We found that consumers render evaluatively different judgments of the same product, depending on which attribute is activated by contextual factors. The present research suggests that the two streams of research can be integrated fruitfully.

This study also has an interesting implication for research on persuasion. The findings suggest an indirect persuasive attempt in which one provides seemingly neutral information (e.g., weight of a bag) and primes consumers to encode the information in terms of the target benefit (e.g., ease of handling). Such indirect approaches to persuasion are likely to offer several advantages over traditional techniques claiming the target benefit directly. For example, indirect persuasion has been found to generate less negative cognitive responses and to be more stable over time, compared with direct persuasion (Yi 1990a, 1990c).

Finally, this study provides useful insights into the process of priming effects by employing step-down analyses with structural equation models. Such analyses provide better understanding of priming effects by incorporating the relationships among dependent variables and correcting for measurement error (Bagozzi and Yi 1989). In contrast, existing research either used a single dependent variable or relied on MANOVA or separate ANOVAs of observed variables (e.g., Soldow and Principe 1981). For example, the univariate test for PI in the present study may have been interpreted as evidence for the direct link from priming to PI. However, step-down analyses ruled out this possibility by indicating that the priming effect on PI does not hold unless Ab is considered as a mediating variable.

The present research is also relevant to practitioners of advertising. First, an ad context can either inhibit or facilitate the effects of a particular ad on brand evaluations. The specific attributes (e.g., ease of use or versatility) relevant to evaluating an advertising product may vary in salience as a function of its context, and these variations may influence the favorability of brand evaluation. By showing that an ad context is not just a benign background but can influence the effectiveness of an ad, this study expands the scope of both strategic and tactical approaches to persuasion.

Also, the present study helps advertisers to understand the unintended effects of the ad context. If the ad context primes negative interpretations of the product, perceptions of the advertised product will be negatively affected. One should avoid placing the ad in such an environment. Alternatively, one should proactively create an advertising environment that can enhance the effect of the target ad. Finally, this study provides a new perspective into the effects of competitive advertising (cf. Burke and Srull 1988). The study suggests the possibility that ads for competing brands might be beneficial if competing brands can prime certain product attributes which are relevant to interpreting information about the target brand.

Several limitations of the present study are in order. First, this study used somewhat strong priming manipulations in a lab experiment, and one might argue that priming is unlikely to occur in the actual advertising context. However, advertisements for products that use product features or benefits as a basis for positioning (e.g., Budget Rent-A-Car) may prime certain attributes (e.g., economy) to consumers. Also, magazine articles (e.g., crime story) may make certain attributes (e.g., safety) salient to consumers. One should assess the extent to which the general ad context primes product attributes and examine whether the findings are generalizable.

It should also be noted that measurement procedures for attribute accessibility have been intrusive. A comparison of the results from Experiments 1 and 2 provides some evidence for the biasing effects of the attribute elicitation task on brand evaluations. One needs to develop subtler ways to assess attribute accessibility, which would not bias subjects' processing of ad information. This study focused on cognitive priming effects of the ad context, but it should be mentioned that the ad context can also have affective priming effects (Erdley and D'Agostino 1988). For example, a magazine article or a TV program may evoke certain affective/feeling states temporarily, such as when it contains pleasant or unpleasant stories. Indeed, Yi (199Ob) shows that an ad context can prime affective reactions among ad recipients and influence their attitude toward the ad, which in turn affects brand evaluations.

We have examined competitive ads as a contextual priming cue that affects the accessibility of product attributes in brand evaluations. Future research can focus on other factors (e.g., point-of-purchase stimuli or program contexts) that may prime certain attributes to consumers in processing product information. Future research should also investigate variables that moderate the contextual priming effects such as involvement and knowledge.

REFERENCES

Bagozzi, Richard P. and Youjae Yi (1989), "On the Use of Structural Equation Models in Experimental Designs," Journal of Marketing Research, 26 (August), 271-284.

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, 14 (September), 141-154.

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

Bray, James H. and Scott E. Maxwell (1982), Multivariate Analysis of Variance, Beverly Hills, CA: Sage.

Burke, Raymond R. and Thomas K. Srull (1988), "Competitive Interference and Consumer Memory for Advertising," Journal of Consumer Research, 15 (June), 55-68.

Chook, Paul H. (1985), "A Continuing Study of Magazine Environment, Frequency, and Advertising Performance," Journal of Advertising Research, 25 (4), 23-33.

Erdley, Cynthia A. and Paul D'Agostino (1988), "Cognitive and Affective Components of Automatic Priming Effects," Journal of Personal and Social Psychology, 54 (5), 741-747.

Gardner, Meryl Paula (1983), "Advertising Effects on Attributes Recalled and Criteria Used for Brand Evaluations," Journal of Consumer Research, 10 (December), 310-318.

Helson, Harry (1964), Adaptation-Level Theory, New York: Harper and Row.

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

Higgins, E. Tory, John A. Bargh, and Wendy Lombardi (1985), "Nature of Priming Effects on Categorization," Journal of Experimental Psychology: Learning, Memory, and Cognition, 11 (January), 59-69.

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 Kihlstrom, Hillside, NJ: Erlbaum, 69-122.

Jaccard, James, David Brinberg, and Lee Ackerman (1986), "Assessing Attribute Importance: A Comparison of Six Methods." Journal of Consumer Research, 12 (March), 463-468.

Kisielius, Jolita and Brian Sternthal (1986), "Examining the Vividness Controversy: An Availability-Valence Interpretation," Journal of Consumer Research, 12 (March), 418-431.

MacKenzie, Scott B., Richard J. Lutz, and George E. Belch (1986), "The Role of Attitude Toward the Ad as a Mediator of Advertising Effectiveness: A Test of Competing Explanations," Journal of Marketing Research, 23 (May), 130-143.

Meyers-Levy, Joan (1989), "Priming Effects on Product Judgments: A Hemispheric Interpretation," Journal of Consumer Research, 16 (June), 76-86.

Payne, John W. (1982), "Contingent Decision Behavior," Psychological Bulletin, 92 (2), 382-402.

Roy, J. (1958), "Step Down Procedure in Multivariate Analysis," Annals of Mathematical Statistics, 29 (December), 1177-1187.

Ryan, Michael J. and Morris B. Holbrook (1982), "Importance, Elicitation Order, and Expectancy x Value," Journal of Business Research, 10 (September), 309-317.

Shavitt, Sharon and Russell H. Fazio (1990), "Effects of Attribute Salience on the Consistency of Product Evaluations and Purchase Predictions," in Advances in Consumer Research, Vol. 17, eds. Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, Provo, UT: Association for Consumer Research, 91-97.

Soldow, Gary F. and Victor Principe (1981), "Response to Commercials as a Function of Program Context," Journal of Advertising Research, 21 (2), 59-65.

Srull, Thomas K. and Robert S. Wyer (1980), "Category Accessibility and Social Perception: Some Implications for the Study of Person Memory and Interpersonal Judgments," Journal of Personality and Social Psychology, 38 (June), 841 -856.

Tversky, Amos and Daniel Kahneman (1973), "Availability: A Heuristic for Judging Frequency and Probability," Cognitive Psychology, 5 (September), 207-232.

Wyer, Robert S. and Thomas K. Srull (1981), "Category Accessibility: Some Theoretical and Empirical Issues Concerning the Processing of Social Stimulus Information," in Social Cognition: The Ontario Symposium, eds. E. Tory Higgins, C. P. Herman, and M. P. Zanna, Hillsdale, N.J.: Erlbaum, 161-197.

Wyer, Robert S. (1986), "Human Cognition in Its Social Context," Psychological Review, 93 (July), 322359.

Yi, Youjae (199Oa), "The Indirect Effects of Advertisements Designed to Change Product Attribute Beliefs," Psychology and Marketing, 7 (Spring), 47-63.

Yi, Youjae (199Ob), "Cognitive and Affective Priming Effects of the Context for Print Advertisements," Journal of Advertising, 19 (2), 40-48.

Yi, Youjae (1990c), "Direct and Indirect Approaches to Advertising Persuasion: Which is More Effective?" Journal of Business Research, 20 (June), 279-29 1.

Yi, Youjae (199Od), "The Effects of Contextual Priming in Print Advertisements," Journal of Consumer Research, 17 (September), 215-222.

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