Effects of Advertising and Involvement on Attitude Accessibility: a Cross-Cultural Replication and Extension

ABSTRACT - This research extended Berger and Mitchell's (1989) study on the effects of ad repetition on the non-evaluative dimensions of attitude. Involvement was incorporated as an added moderating factor and the study was conducted with Singaporean subjects. Results partially replicated the original findings. However, involvement was not found to impact attitude accessibility. Implications of the results are discussed and directions for future research furnished.


Siew Meng Leong, Swee Hoon Ang, and Liang Sze Tan (1994) ,"Effects of Advertising and Involvement on Attitude Accessibility: a Cross-Cultural Replication and Extension", in AP - Asia Pacific Advances in Consumer Research Volume 1, eds. Joseph A. Cote and Siew Meng Leong, Provo, UT : Association for Consumer Research, Pages: 257-260.

Asia Pacific Advances in Consumer Research Volume 1, 1994      Pages 257-260


Siew Meng Leong, National University of Singapore

Swee Hoon Ang, National University of Singapore

Liang Sze Tan, Union Bank of Switzerland


This research extended Berger and Mitchell's (1989) study on the effects of ad repetition on the non-evaluative dimensions of attitude. Involvement was incorporated as an added moderating factor and the study was conducted with Singaporean subjects. Results partially replicated the original findings. However, involvement was not found to impact attitude accessibility. Implications of the results are discussed and directions for future research furnished.


Recent theorizing in social psychology has delineated several dimensions of attitude (cf. Zanna and Fazio 1982). Such accounts have primarily distinguished its evaluative and non-evaluative aspects. Specifically, such non-evaluative attitudinal dimensions as attitude accessibility and attitude confidence have been examined as possible moderators of the attitude-behavior relationship (Fazio, Powell, and Williams 1989).

Consumer research has traditionally focused on the evaluative dimension of attitude, while neglecting its non-evaluative aspects. Clearly, as in social psychology, this lack of attention needs to be addressed for further theoretical progress on the attitude construct to be achieved. In particular, it may be instructive to explore whether advertising and involvement affect the non-evaluative dimensions of consumer brand attitudes. In this vein, Berger and Mitchell (1989) tested the possibility that ad repetition impacted attitude accessibility and attitude confidence. Their results indicated that attitudes formed on the basis of ad repetition are similar to those formed from direct experience in that they are more accessible from memory, are held with greater confidence, and are more predictive of subsequent behavior than those based on a single ad exposure.

Berger and Mitchell's (1989) study represents an important initial attempt which challenges the broadly cherished assumption that advertising must first change attitudes to change behavior. It provides empirical evidence suggesting the possibility that advertising, or, more specifically, its repetition, may influence behavior while having a minimal effect on brand evaluations (cf. Smith and Swinyard 1983).

However, as with much of consumer research, additional work needs to be done to augment the conceptual and empirical foundations of the Berger and Mitchell (1989) study. This work is particularly salient for replication and extension as it represents the discipline's initial attempt to systematically explore the non-evaluative aspects of consumer attitude. We attempt to do so in this research by considering a second factor which may impact the non-evaluative dimension of attitude-the involvement level of consumers. Involvement has been a major focus of much research in consumer behavior (cf. Zaichowsky 1985). It would therefore be natural and appropriate to extend Berger and Mitchell's (1989) work by incorporating this factor explicitly in an empirical study. After all, involvement has been shown to moderate the evaluative aspects of consumer attitudes in past research (e.g., Petty, Cacioppo, and Schumann 1983).

A second direction of extending Berger and Mitchell's (1989) study relates to adding a cross-cultural dimension. In particular, as with much of consumer research, North American subjects were employed in their experiment. It would be logical to assess whether their findings are culture-bound by using subjects from a different continent and background. Once again, only the evaluative aspects of attitude have been studied across different cultures. For example, Lee and Green (1991) provided a cross-cultural examination of the Fishbein Behavioral Intentions Model with a sample of South Korean respondents. Would the North American findings pertaining to non-evaluative attitude dimensions be similarly replicated in another cultural setting?

This study thus attempts to replicate and extend Berger and Mitchell's (1989) work on the non-evaluative aspects of attitude by analyzing the impact of advertising repetition and involvement on attitude accessibility using a sample of Singaporean subjects. In so doing, it hopes to furnish additional conceptual and empirical insights into the attitude construct. In the remainder of this paper, we first furnish a review of the literature on attitude accessibility and involvement with a view towards generating the hypotheses to be tested in this research. The research method employed in our study is then detailed followed by the results obtained from it. Implications of our findings are then given and directions for future research suggested.


Attitude Accessibility

Attitude accessibility is the response latency to an attitudinal enquiry (Fazio 1986). Presumably, the speed at which people respond to direct inquiries about their attitudes covaries with the likelihood of spontaneous activation of attitudes upon the mere observation of an attitude object. Consequently, it has been postulated that the stronger the object-evaluation association, the more accessible the attitude was assumed to be, as reflected by faster latency responses to attitude enquiries.

Berger and Mitchell (1989) found that attitudes based on indirect experience may be made just as accessible as those based on direct experience, given certain conditions. Their results showed that under high involvement, ad repetition results in attitudes that are just as accessible as those formed under direct experience. Hence, replicating Berger and Mitchell's (1989) study, H1 to H3 state:

H1:Attitudes are less accessible under single exposure than under four exposures.

H2:There is no difference in attitude accessibility under four exposures from that under direct experience.

H3:Attitudes are less accessible under single exposure than under direct experience.


Kardes (1988) suggested that as involvement increases, the amount of effort expended in information processing increases. It has also been argued that people have more extensively developed "schema" or cognitive structures for self-relevant information, thus facilitating the processing of information (Markus 1977). With this increased cognitive effort allocated to information processing under high involvement conditions, object-evaluation associations will be stronger and the attitude will be more accessible on mere observation of an attitude object (Fazio 1986). Conversely, in low involvement, messages received are of little personal relevance and so less cognitive effort will be involved in processing information. This results in lower attitude accessibility. Therefore, H4 states:

H4:Relative to low involvement situation, high involvement results in more accessible attitudes.



A 3x2 between-subjects factorial design was used. The factors were ad exposure (single ad exposure versus four ad exposures versus direct experience) and involvement (high versus low). Subjects were 107 college and high school students, randomly assigned to the treatment cells. The attitude object chosen was chocolates. This product met such criteria as availability to subjects, identifiability by subjects, wide range and selection, and sampling potential in a lab setting. Following pretesting, four fictitious brand names were used: Choco-mint, Chocolate, Wafer Bar, and Swiss-li. All were vividly descriptive of the chocolates chosen based on the thoughts and feelings elicited from pretest subjects.


Ad Exposure. In the single-exposure condition, subjects were shown the four print ads, one for each brand. The ads were standardized in structure. They included a large headline which identified the brand, a picture of the chocolate on its package, a subheadline describing its expected taste, and the ingredients that went into the chocolate. In the four-exposure condition, the ads were randomized within four blocks, each block containing four ads. The order in which the four ads were placed in each block was also randomized to minimize possible order effects. Subjects saw each ad for 15 seconds with no flipping back to the previous ad. In the direct experience condition, subjects tasted the chocolates.

Involvement. High and low involvement were manipulated following the method employed by Apsler and Sears (1968). Subjects in both groups received the same communication but high-involvement subjects were led to believe that the issue had a direct impact on or personal relevance to them. Specifically, high-involvement subjects were told that they would be evaluating several brands of chocolates to help formulate the marketing strategy for these products when they were launched in their country in two months' time. Those in the low-involvement condition were told that the session was to gauge their response to various chocolates that would be introduced in a neighboring country in a year's time. Seven seven-point items (a=.94) were used to measure the involvement construct (Zaichkowsky 1985). Pretest results showed differences in involvement between the high and low involvement groups were in the right direction (t=5.08, p<.01).

Dependent Variable

A computer program was developed to measure attitude accessibility. Some 28 questions relating to the four brands of chocolates were flashed on the computer screen randomly. Twenty items were attitudinal in nature (e.g., "Is Choco-mint good?"), while the other eight were attribute questions (e.g., "Is Choco-mint peppermint flavored?"). Subjects indicated 'yes' or 'no' to the questions by pressing one of the labelled shift keys on the typing pad. The program measured and recorded the response time in milliseconds as well as the response. The response time was that between the appearance of the question on screen and when the subject pressed one of the keys. Each question was separated by a two-second interval and a "get ready" signal before the next question appeared. If no answer was provided after 25 seconds, the next question was automatically shown and a time of zero was recorded. This measurement procedure is similar to that employed by Fazio et al. (1982) and Berger and Mitchell (1989).

Attitude accessibility was then measured by response latency scores based on 20 evaluative items, five for each brand of chocolate. Coefficient alphas were obtained in excess of .50 for the latency measures. Possibly, these weaker reliabilities may be due to the "yes/no" responses being used to represent attitudes after translation to latency indicants of attitude accessibility. The five response latencies for each brand were averaged to yield a mean attitude accessibility score for each brand labelled TIMEA, TIMEB, TIMEC, and TIMED for Choco-mint, Chocolate, Wafer Bar, and Swiss-Li respectively.


Subjects were run in groups not exceeding six. They were greeted upon entering the computer lab used for the experiment and asked to sit in front of a personal computer already booted up. They were requested not to touch the terminal but to listen carefully while a script inducing the involvement condition was read to them. Those in the direct experience condition were provided bite-sized pieces of chocolates for each brand. Sample bars of chocolates were placed beside the bite-sized pieces and glasses of water furnished for drinking between tasting.

Following this, filler tasks were given to clear short-term memory. Subjects were then given instructions for the attitude accessibility test. Subjects were given six questions to familiarize themselves with the terminal. The session concluded with the collection of demographic information and debriefing.


Descriptive statistics for the various experimental conditions are shown in the Table.

H1: H1 stated that attitude is less accessible under single exposure than under four exposures. H1 was supported for the Swiss-Li brand as TIMED was greater in the multiple than single exposure condition (t=1.83, p<.05), and marginally for TIMEA and TIMEC for the Choco-mint and Wafer Bar brands (t's>1.27, p's<.10), but not for TIMEB for the Chocolate brand (t=1.11, p>.10).

H2: H2 stated that there is no difference in attitude accessibility under the multiple exposure and direct experience conditions. H2 was supported by TIMEC and TIMED for the Wafer Bar and Swiss-Li brands (t's<.75, p's>.10), and marginally for TIMEA and TIMEB for the Choco-mint and Chocolate brands (t's>1.79, p's<.10).

H3: H3 stated that attitudes are less accessible under single exposure than direct experience. This was not supported by the results of all four brands (t's<.36, p's>.10).

H4: H4 stated that under high involvement, attitudes are more accessible than under conditions of low involvement. H4 was not supported as the differences in accessibility times were not significant across the two levels of involvement for all four brands (t's<.35, p's>.10).


The results obtained from Hypotheses 1 to 3 partially replicate those by Berger and Mitchell (1989). Indeed, similar findings were obtained when the analyses for these hypotheses were conducted only for subjects in the high-involvement condition. However, the magnitude of differences obtained was not as large as those in previous research (Fazio et al. 1982; Fazio, Herr, and Olney 1983; Fazio, Powell, and Herr 1983). One explanation may be that subjects were not made to explicitly express their attitudes after each exposure to the ad. Instead, it may be inferred that after four exposures, subjects have activated their attitudes a sufficient number of times to allow easy accessibility of their attitudes. This is reflected by the finding that four exposures result in equally accessible, or even more accessible attitudes, as compared with direct experience with the attitude object.



Managerially, this finding implies that wherever possible, marketers should allow potential customers to try their product, either by way of sampling or product demonstration. For products, particularly capital intensive ones, where sampling or demonstration is not feasible, ad repetition appears to be a viable alternative.

An interesting finding pertained to the fact that attitude accessibility was different for the four brands of chocolates investigated insofar as H1 and H2 were concerned. This suggests that attitude accessibility may be brand-specific. The present data does not provide a sufficient basis to ascertain why this may be the case. The chosen brands were not identifiable by pretest subjects based on taste. Possibly, other cues which may have impacted attitude accessibility were not controlled for. Fazio, Powell, and Williams (1989) showed that a momentarily salient feature (whether they were positioned in the front or back row) of attitude objects that were low in accessibility had a greater impact on consumer choice of chocolates than those high in accessibility. As their research focused on selection behavior, added work is necessary to ascertain whether such effects also hold for non-evaluative aspects of attitude.

Involvement was found to have no effect on attitude accessibility. The most likely explanation of this may be discerned from the manipulation check scores for the high-involvement subjects which averaged 29.3 in the pretest. This is not significantly different from the scale midpoint of 28. Essentially, they may have been only moderately involved with the experimental task. Hence, future research may fine-tune the manipulation of involvement to assess its effect on attitude accessibility. Alternative approaches to inducing subject involvement may also be employed.

In general, it appears that North American results on the non-evaluative dimensions of attitude are replicable cross-culturally. Hence, the statistical conclusion validity of the original findings is enhanced, suggesting that they are not likely to be due to Type I error. This parallels Lee and Green's (1991) results indicating that the Fishbein Behavioral Intentions Model can be extended to the Asian domain. Nonetheless, only Singaporean subjects were employed to assess the robustness of the results in this study. Moreover, as the present inferences are based on between-studies comparisons, a more direct test using data collected from both Singaporean and American samples appears warranted. To further buttress the empirical generality of these results, subjects from other countries and cultures need also be used in future work. Beyond attitude accessibility investigated here, such other aspects as attitude confidence should be assessed for their cross-cultural generality. Indeed, such exploration may be extended to include replicating Berger and Mitchell's (1989) results on the subsequent impact of these factors on the attitude-behavior relationship.

Collectively, the agenda for future research on the non-evaluative aspects of attitude offers much promise for consumer researchers. Aside from augmenting replication studies so lacking in the field (cf. Monroe 1991), they may contribute more substantively towards a better understanding of the attitude construct as well.


Apsler, R. and D. O. Sears (1968), "Personality, Involvement, and Attitude Change," Journal of Personality and Social Psychology, 9, 162-168.

Berger, Ida E. and Andrew A. Mitchell (1989), "The Effect of Advertising on Attitude Accessibility, Attitude Confidence, and the Attitude-Behavior Relationship," Journal of Consumer Research, 16 (December), 269-279.

Fazio, Russell H. (1986), "How Do Attitudes Guide Behavior?" in Handbook of Motivation and Cognition: Foundations of Social Behavior, eds. Richard M. Sorrentino and E. Tory Higgins, New York: Guildford, 204-243.

Fazio, Russell H., Jean-Mei Chen, Elizabeth C. McDonel, and Steven J. Sherman (1982), "Attitude Accessibility, Attitude-Behavior Consistency, and the Strength of the Object-Evaluation Association," Journal of Experimental Social Psychology, 18 (4), 339-357.

Fazio, Russell H., Paul M. Herr, and Timothy J. Olney (1984), "Attitude Accessibility Following a Self-Persuasion Process," Journal of Personality and Social Psychology, 47 (2), 277-286.

Fazio, Russell H., Martha C. Powell, and Paul M. Herr (1983), "Toward a Process Model of Attitude-Behavior Relation: Accessing One's Attitude Upon Mere Observation of the Attitude Object," Journal of Personality and Social Psychology, 44 (4), 723-735.

Fazio, Russell H., Martha C. Powell, and Carol J. Williams (1989), "The Role of Attitude Accessibility in the Attitude-to-Behavior Process," Journal of Consumer Research, 16 (December), 280-288.

Kardes, Frank R. (1988), "Spontaneous Inference Processes in Advertising: The Effects of Conclusion Omission and Involvement on Persuasion," Journal of Consumer Research, 15 (September), 225-233.

Lee, Chol and Robert T. Green (1991), "Cross-Cultural Examination of the Fishbein Behavioral Intentions Model," Journal of International Business Studies, 22 (2), 289-305.

Markus, Hazel (1977), "Self-Schemata and Processing Information About the Self," Journal of Personality and Social Psychology, 35, 63-78.

Monroe, Kent B. (1991), "Editorial: On Replications in Consumer Research: Part I," Journal of Consumer Research, 18 (June), i-ii.

Petty, Richard E., John T. Caccioppo, and David Schumann (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of Consumer Research, 10 (September), 135-146.

Smith, Robert E. and William R. Swinyard (1983), "Attitude-Behavior Consistency: The Impact of Product Trial Versus Advertising," Journal of Marketing Research, 20 (August), 60-67.

Zaichowsky, Judith Lynne (1985), "Measuring the Involvement Construct," Journal of Consumer Research, 12 (December), 341-352.

Zanna, Mark P. and Russell H. Fazio (1982), "The Attitude-Behavior Relation: Moving Toward a Third Generation of Research," in Consistency in Social Behavior: The Ontario Symposium, Vol. 2, eds. Mark P. Zanna et al., Hillsdale, NJ: Lawrence Erlbaum, 283-301.



Siew Meng Leong, National University of Singapore
Swee Hoon Ang, National University of Singapore
Liang Sze Tan, Union Bank of Switzerland


AP - Asia Pacific Advances in Consumer Research Volume 1 | 1994

Share Proceeding

Featured papers

See More


Social Sharing of Negative Emotions in Virtual Travel Communities

Clara Koetz, Rennes School of Business
Anke Piepenbrink, Rennes School of Business

Read More


Too Much of a Good Thing? Consumer Response to Changes in Brand Essence

Tarje Gaustad, Kristiania University College
Bendik Samuelsen, BI Norwegian Business School
Luk Warlop, Norwegian School of Management, Norway
Gavan Fitzsimons, Duke University, USA

Read More


Paying to Purchase a Conversation Topic

Hillary Wiener, University at Albany
Joshua Wiener, Oklahoma State University, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.