The Impact of Repetition on Advertisement Misconprehension and Effectiveness

Mark I. Alpert, The University of Texas at Austin
Linda L. Golden, The University of Texas at Austin
Wayne D. Hoyer, The University of Texas at Austin
ABSTRACT - Substantial interest has recently arisen in the study of advertisement miscomprehension and the role of repetitions in that inquiry . This study extends previous work in this area by specifically varying the number of exposures to advertisement messages for four products and measuring the impact on miscomprehension and other measures of advertisement effectiveness variables. Our findings suggest that repetition may not improve comprehension, and there is some evidence of a wearout effect. Problems and opportunities for further research are also addressed.
[ to cite ]:
Mark I. Alpert, Linda L. Golden, and Wayne D. Hoyer (1983) ,"The Impact of Repetition on Advertisement Misconprehension and Effectiveness", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 130-135.

Advances in Consumer Research Volume 10, 1983      Pages 130-135

THE IMPACT OF REPETITION ON ADVERTISEMENT MISCONPREHENSION AND EFFECTIVENESS

Mark I. Alpert, The University of Texas at Austin

Linda L. Golden, The University of Texas at Austin

Wayne D. Hoyer, The University of Texas at Austin

ABSTRACT -

Substantial interest has recently arisen in the study of advertisement miscomprehension and the role of repetitions in that inquiry . This study extends previous work in this area by specifically varying the number of exposures to advertisement messages for four products and measuring the impact on miscomprehension and other measures of advertisement effectiveness variables. Our findings suggest that repetition may not improve comprehension, and there is some evidence of a wearout effect. Problems and opportunities for further research are also addressed.

INTRODUCTION

The concept of message comprehension and miscomprehension is recognized in several test models of persuasive communication (e.g., McGuire 1969 Engel and Blackwell 1989). Recent interest in examinations of the nature and extent of miscomprehension of advertising (vs. nonadvertising) messages has been shown to be substantial. A study by Jacoby, Hoyer, and Sheluga (1980) has attracted a great deal of attention and some controversy. One thing that supporters and critics of the study's methods and findings agree upon is the importance of research addressing itself to miscomprehension or advertising messages. No doubt it is partially due to the important implications for public policy that this study has generated the controversy that it has. As a result, several important conceptual and methodological issues in defining and studying miscomprehension have become salient. The purpose or the present study is to address directly one of these key issues--the effect of repetition on advertising miscomprehension. In addition, we shall also examine theoretical and empirical issues relevant to inquiry concerning repetition, comprehension, and other measures of advertising effectiveness.

Previous Findings on Miscomprehension of Advertising

Tile first study examining the miscomprehension of advertising was conducted by Jacoby and Hoyer (1982). In this study, the miscomprehension rates of sixty television stimuli were assessed in a forced exposure laboratory setting by means of six item multiple choice quizzes for each stimulus. The main purpose was two-fold: (l) to estimate the e :tent of miscomprehension regarding televised messages, and (9) to compare miscomprehension rates across different types of messages (e.g., advertisements, program content, public service announcements). The major findings were that the approximate rate of miscomprehension was 29.6 percent and that there were few differences across different categories of stimuli. Since this figure has been taken (sometimes out of context) as a "benchmark" miscomprehension figure for television communications that are not inherently deceptive, it has been the center or considerable attention. Specifically, this figure has been questioned regarding whether it is an accurate estimate of the true miscomprehension rate. The validity of measuring miscomprehension after a single exposure in a laboratory setting has been questioned (Mizerski 1981). Since most consumers are exposed to the same advertisement many times, the effect of repetition on miscomprehension is an important question.

Previous Research on Repetition's Effects on Miscomprehension and Other Advertising Variables

With respect to repeated exposures to an advertisement and the impact on miscomprehension, it has been suggested that multiple exposures should lower the miscomprehension rate. According to Mitchell and Olson (1976), the consumer may not be able or motivated to process the entire message, due to situational distractions or message complexities. Thus, it may take several repetitions of the message before information is completely understood. Indeed, numerous studies -(e.g., Sawyer 1973, Zielske 1959) have indicated that the effects of a persuasive message are strengthened by repetitions up to a point. As a result has been suggested that the miscomprehension rates evidenced by the Jacoby and Hoyer study (1982) are overestimates of the true miscomprehension rate in a typical advertising campaign.

Differences in the methodologies employed in the literature of repetition effects on advertising variables make it difficult to generalize regarding the nature or its impact. Studies have varied from eight repetitions of the same ad (with buffers) in less than six minutes (Sawyer 1973) to varying repetitions with varying numbers of executions of similar (or identical) messages over varying time periods up to several weeks (Calder & Sternthal 1980 ). For example, Zielske (1959) presented data that indicated that concentrated repetition (as is typical in many laboratory studies) led to a dramatic rise in recall, but recall declined rapidly upon removal of the stimulus. A spaced exposure, on the other hand, led to a fairly constant recall rate over time. In addition, the extent to which attention was forced," voluntary, or used advertisements embedded (and sometimes disguised) within program content also varies in the literature.

In light of the difficulties in comparing repetition studies, two major theories of the effects of repetition can nevertheless be identified.

The first theory of repetition effects essentially adopts an information processing perspective. The major thesis is that individuals are not able to process and store all of the product information in memory based on one exposure. Rather, multiple exposures to the stimulus are needed to produce several major effects (Mitchell and Olson 1976): (l) Information is rehearsed and stored in memory, (2) links between brands and attributes are strengthened, (3) positively evaluated briefs will result in a change in attitude, and (4) continued repetition of the stimulus inhibits the forgetting of this belief. It must be noted, however, that these effects will be present only up to a point (usually around four repetitions), after which wearout occurs. That is, at a certain point individuals have lean ed the information and not only is there no increment in learning, but a decline in the affective rating of the ad. Thus, people apparently become bored with the message and the end result is increasing negative effect.

Several major predictions regarding the effects of repetition on miscomprehension can be suggested on the basis of this review. Basically, it can be posited that repetition will have a facilitating effect on comprehension up until the message has been completely processed (possibly after four repetitions). After this point, repetition should have either no effect or a negative effect on comprehension (due to the negative affect created).

An alternative theory is based on the assumption that advertisements are basically low involvement stimuli (Krugman, 1965; Ray et al. 1973; Rothschild 1979). That is, due to the fact that ads are not great in personal relevance, consumers are not motivated to process the information contained in the ad. Rather, only global impressions or key points are stored in memory. Further, based on the notion that repetition enhances liking (e.g., Zajonc 1968), it is suggested that repetition of a simple message will breed familiarity and liking of the product over time.

The adoption of this perspective would result in different predictions of repetition effects on miscomprehension. The notion that consumers are not motivated to process the ad in any great detail would lead to the hypothesis that miscomprehension of detailed aspects of the ad should be fairly substantial and should not be affected by any repetition of the advertisement. Rather, only global, broad level information (brand name and affect) are stored. Further, an increase in attitude toward the ad should be in evidence as repetitions increase.

Summary and Hypotheses

In summary, the purpose of the present study is to examine the effects of repetition on advertising miscomprehension. Due to the absence of previous research on this topic and to the presence of competing theories of repetition effects, hypotheses are presented in the form or competing exploratory propositions.

Hypothesis 1 -- Repetition and Miscomprehension

H1a: Repetition will decrease miscomprehension up to a point in repetitions, after which these effects will subside. (information processing)

H1b: Repetition will have no effect on miscomprehension due to the low motivation to process the ad. (low involvement hypothesis)

Hypothesis 2 -- Repetition and Affect

H2a: Repetition will cause an increase in affect with number or exposures, then a decrease in affect toward the ad. (information processing hypothesis)

H2b: Repetition will cause an increase in affect toward the ad over time. (low-involvement hypothesis)

Hypothesis 3 -- Repetition and Purchase Intention

H3a: The relationship between purchase intentions and repetitions will be curvilinear. That is, purchase intention will first increase and then decrease. (information processing hypothesis)

H3b: Repetition will result in an increase (linear) in intentions over time. (low involvement hypothesis)

Hypothesis 4 -- Repetition and Perceived Quantity or Information

H4a: As habituation occurs, the perceived quantity or information will decrease with repeated exposures. That is, there will be a curvilinear relationship such that perceived quantity or information will first increase and then decrease (information processing hypothesis)

H4b: Repetition will have no effect on perceived quantity of information due to low processing motivation. (low-involvement processing)

Hypothesis 5 -- Repetition and Believability

H5: Repetition should have no effect on believability of the advertising message. (both theories)

METHODOLOGY

This study replicates much of the methodology used by Jacoby and Hoyer (1989) and extends it to examine the effects of repetitions. Although there are differences between the two studies (sample characteristics and administration format), the methodological similarities do allow for comparison or one-exposure miscomprehension and repeated exposures miscomprehension.

Advertisement Selection

Four thirty-second television advertisements that were included in the Jacoby and Hoyer study (1982) were selected as experimental stimuli. These advertisements were selected on the basis of three criteria. First, since the main interest of the present study was on the effect of repetition, it was necessary to use advertisements that have never been previously viewed. This was necessary because any exposures to the ad prior to the experiment would have made the hypothesis less clear. Several of the advertisements obtained from the Jacoby/Hoyer stimulus capes were acquired (from the Midwest) and thus these ads were unfamiliar to the Southwestern audience. Second, it was desirable to include stimuli that covered the range or miscomprehension levels (i.e., low to high). Thus, based on the normative races evidenced in the Jacoby and Hoyer (1989) study, the ads low, medium, and high in miscomprehension were selected. Finally, only product categories that would be reasonabLy purchased by a student sample were felt appropriate for inclusion.

Previous research indicated a pattern of satiation with repetition, in which attention is maximized at two to four exposures and is followed by a decline as the number or exposures increases (Grass 1968). Five exposures was expected to provide ample opportunity to observe the effects of miscomprehension over time.

In order to implement the study a seating chart was made of the classroom where the mass-section marketing class was held. Students were instructed to remain in the same seats and a packet of questionnaires was assigned to each seat. The students were told that they were participating in a study of the effects of mood on perception or advertisements.

Each packet contained four placebo questionnaire booklets and one experimental treatment questionnaire booklet. The repetition level assigned to a particular seat determined which questionnaire was the experimental treatment questionnaire. Every sixth seat had the same questionnaire order in the packet.

At any one time of administration twenty percent or the subjects were answering the miscomprehension and other dependent variable questions. The remaining eighty percent of the subjects were answering placebo questionnaires. For example, students in the single exposure group answered the miscomprehension and other dependent variable questions immediately following the first exposure and responded to placebo questions on all subsequent exposures, while subjects in the fifth exposure treatment filled out four placebo questionnaires before answering the miscomprehension (and other dependent variable) questions on the fifth administration.

At each administration, subjects took the top questionnaire from their packet, making sure it had the appropriate administration time in the upper right hand corner (one, two, three, four, or five). It should be noted that each version had an identical top sheet (mood questions) so that subjects would be unaware that each questionnaire (booklet) contained different questions.

At the end of each administration, the completed treatment level was placed in the back of the packet, so the next appropriate questionnaire would be facing the front. Because there were several cross-checks built into the administration system, the administration proceeded smoothly with subjects easily returning to the same packet and the appropriate questionnaire for each set of exposures.

Questionnaire Content

The first section of the questionnaire contained six snort questions about the subjects' moods, which were answered on a five-point scale with extremes labeled "strongly agree' (5 ) and "strongly disagree" (1). The ma j or purpose of these questions was to lend credibility to the cover story (the effect of mood on responses to advertising). Further, the mood questions were identical in ail forms and administrations. In the experimental treatment, subjects were given a series of instruments that were designed to assess miscomprehension rates. these instruments used the wording employed in the Jacoby and Hoyer study (19&2). Each questionnaire contained six statements and for each statement, subjects were to indicate "true" if the statement was directly stated or implied in the advertisement and "false" is the statement was neither stated nor implied in the advertisement. Of the six statements, three were restatements of facts contained in the advertising copy (or misstatements in the "false" ones), while the other three were logical inferences (both correct and incorrect) that could be drawn from the advertisement. These statements enabled construction of three measures of miscomprehension for each ad: "overall miscomprehension," "factual miscomprehension," and "inference miscomprehension."

In addition, four other dependent variables were measured. These were advertisement believability, quantity of information, liking of the advertisement (affect), and purchase intentions. These were measured on a horizontal seven-point sc ale ranging from tow (1) to high (7) with poles appropriate to the specific variable.

The placebo questionnaires were carefully designed to avoid drawing upon any of the information contained in the advertisement. These were questions concerning the relevant product category, which did not require any attention to (or active processing of) the advertisement. For example, after presenting a stimulus advertisement about domestic beer, placebo subjects were asked to rate the quality of beer in several specified countries. The subject was also asked to indicate the name of two other countries that would produce "high quality" beer.

Procedure

For the first part of each administration, subjects completed the mood questionnaires. They were told to sit quietly upon completion and not to turn the page until so instructed. Nest, subjects were exposed to the first advertisement. After viewing the advertisement, they were instructed to turn the page and answer the next set of questions (i.e. either the miscomprehension and advertisement effectiveness items or a placebo questionnaire). This same procedure was then repeated for three additional stimuli.

Analysis

The first phase in the analysis was scoring the miscomprehension questions. Each subject received three miscomprehension scores for each advertisement. The scores are described below.

Overall miscomprehension--For each true statement answered "false" and each "false" statement answered "true", the subject receives one point. A zero indicates that there was no miscomprehension (all questions answered correctly) and a six indicates total miscomprehension (all questions answered incorrectly).

Fact miscomprehension--One true statement and two false statements were derived from direct statements in the advertisement. Subjects received one point for each incorrect answer, such that a zero is a "perfect" comprehension score and three is a total miscomprehension score.

Inference miscomprehension--As with fact miscomprehension, one true statement and two false statements resulted from direct inferences made by the advertisements. Subjects received one point for each incorrect answer. Zero is no miscomprehension and a three is total miscomprehension.

The data were submitted to a four product by five repetition repeated measures analysis of variance. The number of exposures (one to five) was treated as a between-subjects factor and products were a within subjects factor, with subjects as replicates for products. The data were analyzed for each dependent variable separately: overall miscomprehension, fact miscomprehension, inference miscomprehension, believability, quantity of information, affect and purchase intentions. Only subjects who had been in advertisement administrations prior to their experimental treatment questionnaire were included in the sample. The final sample size was one hundred and thirty-seven respondents.

RESULTS

The results for the analysis of variance are presented in the Table. The statistics for interactions are not included because none of these were significant (at even the .25 level). A significant main effect was found for the product advertised (or advertisement) for all seven dependent variables, but the number of repetitions had no significant impact on miscomprehension. In fact, the only variable that the number of repetitions was significantly associated with was affect towards the advertisement (in a manner suggesting "wearout").

The pattern of mean ratings for product and advertisement stimuli varied across dependent variables. For both overall miscomprehension and fact miscomprehension, the beer advertisement had the highest miscomprehension (39% and 62%), while the restaurant advertisement had the lowest (9.5% and 7%). The order of product means was identical for these two dependent variables. In terms of inference miscomprehension, the photographic portrait advertisement was the most miscomprehended (25%) and the restaurant advertisement was the least miscomprehended (12%).

In general, miscomprehension rates were slightly lower for this study than that of the Jacoby, et al (1980). The present weighted average miscomprehension rate of 22.3% is less than the 29.5% previously found for advertising miscomprehension for these four products, a decline which may be in part due to the use of a student sample, who may have perceived the stimuli more accurately than an average sample.

TABLE

ANALYSIS OF VARIANCE FOR REPETITIONS AND PRODUCTS ON ADVERTISEMENT VARIABLES

The important finding is that no significant decline in miscomprehension rate may be said to be due to repeating the ads up to five times. While it may appear from inspection of the group means that miscomprehension rose and then fell after repetitions, the ratio of within roup variance to between group variance is so large that this inference is not warranted statistically. Repetitions had neither systematic nor significant effect on overall miscomprehension nor its components of miscomprehension of facts and inferences.

On the other hand, as mentioned above, there was a weak (omega2 = 2%) but statistically significant effect (a < .04) or repetitions on attitude cowards the advertisements. Variations in advertisement and product were more strongly associated (omega2 = 13%) with advertisement effect, which is not surprising, given the product and ad variations. [Significant effects of repetition on attitude toward advertisements, as well as significant between-product differences in affect are evidence of manipulation check. Clearly different ads were differently liked, and repetition seemed associated with advertisement wearout.] Within the repetition effect, Liking or the advertisement declined after the first exposure, which may be evidence of the "wearout" effect discussed in some of the literature. This finding would also be consistent with Sternthal and Craig's information processing view of repetition effects (1982, p. 281). They hypothesize that beyond a certain point, people have learned the communicated information (indexed here by relatively low miscomprehension races after one exposure). Thus, people may begin to retrieve their own battery of thoughts, many of which may be more negative than the ones contained in the message that is designed to sell them (perhaps as counter-arguing strategy).

The number of repetitions had no other significant effects on the dependent variables, although the product/ advertisement variable did show some interesting effects. For example, looking at the significant main effects of product on the dependent variables in the study, we see the advertisement (restaurant) that was the least miscomprehended was rated as the most believable, containing the most information, the most well-liked, and elicited the highest purchase intentions. The beer advertisement, which was the most miscomprehended overall and factually, was perceived as the least believable and having the least amount of information. Thus, in terms of product rankings, evaluations of believability and amount of information were similar and evaluations of affect and purchase intentions were similar across Products.

DISCUSSION

The major finding of the present study was that repetition had little effect on miscomprehension. In addition, significant main effects of repetition were not in evidence for believability, amount of information or purchase intentions. Further, whatever negative effect repetition may have had on liking for the advertisement, this effect did not carry over to other dependent variables measuring advertisement effectiveness and purchase intention.

Thus, these findings do not consistently support the propositions suggested by either of the theories of repetition. That is, absence of a main effect for repetition was inconsistent with the curvilinear relationship posited by an information processing perspective, and, while the miscomprehension finding is compatible with a low involvement viewpoint, the results concerning affect (i.e., a decrease in affect over time) are not consistent with this theory.

However, the results of the present study contradict the argument that the estimates of miscomprehension obtained by Jacoby and Hoyer (1982) were unusually high because of a single exposure to the stimulus. However, one must still question the extent to which the present study has truly modeled a real-world repetition effect (i.e., the external validity of a forced exposure design).

At present, one possible explanation for the flat miscomprehension function might be that a single-exposure design with forced exposure may heighten comprehension (vs. a single exposure in the field), therefore understanding miscomprehension somewhat (Jacoby and Hoyer 1982). Hence, from this "low" starting point, repetitions cannot move miscomprehension much lower. Alternatively, boredom may have lowered attention to subsequent ads and the questions regarding them, thus inhibiting improvements in miscomprehension scores. Which of these hypotheses (if either) is valid, or even stronger in impact, is not clear at present, and there is room for creative designs that might test the impact of repetitions on miscomprehensions in a relatively non-reactive but naturalistic setting. With more complex material, different audiences, and in other contexts, repetition may reduce miscomprehension. For the present, it seems reasonable to suppose that wearout in this simulated advertising repetition presentation would approximate that of the the external environment, in which people may also be expected to reach an early plateau of comprehension, due in part to unwillingness to attend fully to previously observed messages. One thing we begin to suspect from this study (generalization from the sample permitting) is that repetition in the Jacoby/Hoyer paradigm could not be per se expected to lower miscomprehension rates, in spite of what some critics have suggested they should have done.

If alternatives to forced exposure are sought, care must be taken to avoid errors present in several simple field exposure studies. That is, it may be misleading to correlate number of exposures to advertising with measures of comprehension and affect. The reason is that persons who are said to be exposed to repetitions, due to their measured number of exposures, are in a sense self-selected. Mortality issues imply that they are more interested in the advertisement or product than those less frequently "exposed"; hence, the measure of repetition may easily confound its effects (e.g., "Did they comprehend better due to repetitions, or did they comprehend better because of involvement with the product that made them interested in acquiring more information--or was there some combination of the two?"). If a "natural" setting is sought, exposure level can- be manipulated at the transmission source, but guaranteeing that reception has occurred either lets you in for criticisms of "forced exposure" settings or else may confound repetition with willingness to watch repeatedly specific stimuli.

A promising type of design for the externally valid examination of repetition effects is illustrated by the Calder and Sternthal study of repetition and advertising wearout (1980). It is perhaps the best example of a naturalistic setting for studying advertisement reactions relatively unobtrusively by imbedding the ads in programs, varying repetitions, timing, and similarity of executions. Extending this design to incorporate measures of miscomprehension might provide confirmation of the conclusions suggested in the present exploratory study. It might also be worth further investigating the intriguing possibility that those ads high in miscomprehension might also be lowest in purchase intention and liking. The relationship between advertisement liking and purchase intention can be confounded in several ways. However, if this link begins to generalize, the role of public policy may be less vital, to the extent that miscomprehended ads may be "punished" by the marketplace. The conditions under which highly miscomprehended ads may be ineffective (e.g., if the audience feels deliberately deceived or misled) may warrant further inquiry. Depending on its (attributed) cause, miscomprehension may have dramatically different impact on consumers, public agencies, and advertisers.

REFERENCES

Calder, B. and Sternthal, B. (1980), "Television Commercial Wearout: An Information Processing View," Journal of Marketing Research, 17 ('lay), 1,3-86.

Engel, J. F. and Blackwell, R. D. (1982), Consumer Behavior, Fourth Edition, Hinsdale, IL: The Dryden Press.

Grass, R. C. (1968), "Satiation Effects of Advertising," Proceedings of the 14th Annual Conference of the Advertising Research Foundation, Ne; York: Advertising Research Foundation, 20-28.

Jacoby, J., Hoyer, W. D., and Sheluga, D A. (1980), Miscomprehension of Televised Communications, New York: American Association of Advertising Agencies.

Jacoby, J. and Hoyer, W. D. (1982), "Viewer Miscomprehension of Televised Communication: Selected Findings, Journal of Marketing, (Fall, in press).

Krugman, H. E. (1965), "The Impact of Television Advertising: Learning Without Awareness," Public Opinion Quarterly, 29, 349-56.

McGuire, W. J. (1969), "The Nature of Attitudes and Attitude Change, " in G. Lindzey and E. Aronson (Eds.), The Handbook of Social Psychology, Second Edition, Vol. 3, Reading, Mass.: Addison-Wesley. 136-314.

Mitchell , A. A. and Olson, J. C. (1976), "Cognitive Effects of Advertising Repetition," Working Paper no. 49, College of Business Administration, Pennsylvania State University, October.

Mizerski, R. W. (1981), "Major Problems in 4 A's Pioneering Study of TV Miscomprehension," Marketing News, June 12.

Ray, M. L., Sawyer, A. G., Rothschild, M. L., Heeler, R. 'I., Strong, E. C., and Reed, J. B. (1973), "Marketing Communications and the Hierarchy of Effects," in P. Clarke (Ed.), New Models for Mass Communication Research, Beverly Hills: Sage. 147-76.

Rothschild, M. L. (1979), "Advertising Strategies for High and Low Involvement Situations," in J. Maloney and B. Silverman (Eds.), Attitude Research Plays for High Stakes, Chicago: American Marketing Association, 74-93.

Sawyer, A. G. (1973), "The Effects of Repetition on Refutational and Supportive Advertising Appeals," Journal of Marketing Research, 10 (February), 93-33.

Sternthal, B. and Craig, C. S. (1982), Consumer Behavior, Englewood Cliffs: Prentice Hall.

Zajonc, R. B. (1968), "Attitudinal Effects of Mere Exposure," Journal of Personality and Social Psychology, Monograph, 9 (2, Part 2), 1-28.

Zielske, H. (1959), "The Remembering and Forgetting of Advertising," Journal of Marketing, 23, 239-43.

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