Advances in Consumer Research Volume 8, 1981 Pages 410-413
VIEWER MISCOMPREHENSION OF TELEVISED COMMUNICATION: A BRIEF REPORT OF FINDINGS
Jacob Jacoby, Purdue University
Wayne D. Hoyer, Purdue University
David A. Sheluga, Purdue University
[This paper is a highly condensed version of a monograph describing a recently conducted investigation sponsored by the American Association of Advertising Agencies. The authors acknowledge with gratitude the many constructive contributions made by the rompers of the Academic and Industry Review Committees established by the AAAA to monitor this project from inception to completion. The members of the Academic Review Committee were Professors Jacob Cohen (Hew York University), Stephen Greyser (Harvard University), and William McGuire (Yale University). The Industry Review Committee members were Mrs. Rena Bartos (J. Walter Thompson), Dr. Seymour Beaks (Leo Burnett), Dr. Theodore Dunn (Benton and Bowles) and Dr. Benjamin Lipstein (originally at SSC&B, now at New York University). The full description is provided in Jacoby, Hoyer, and Sheluga (1980).]
[Wayne Hoyer is now with the Department of Marketing, University of Texas at Austin, Austin, Texas. David Sheluga is now with the Quaker Oats Company, Barrington, Illinois.]
Some of the major findings steaming from a large scale investigation of viewer miscomprehension of televised communication are presented. These findings reveal that virtually all viewers miscomprehend to some extent, and that no communication is immune from being miscomprehended. Further, the average amount to which the core content of each communication was miscomprehended was s surprisingly high 30%. Finally, commercials seen to be slightly better comprehended than program content. A few of the implications of these data are also discussed.
Considerable interest has recently been focussed on issues related to consumer miscomprehension of advertising. Consumer advocates, regulatory agencies, and academic researchers have been devoting an increasing amount of attention to the issues of deceptive and/or misleading advertising and their remedies (e.g., corrective advertising, affirmative disclosure statements). They have not, however, addressed the logically prior set of questions concerning the existence, degree, correlates, and causes of miscomprehension. The evidence that is available in these regards tends to be fragmentary, anecdotal, and impressionistic.
Implicit in the arguments advanced by many consumer advocates is the assumption that commercial advertising is prone to causing miscomprehension whereas other forms of communication are generally not so prone. A problem with this view, however, is that it ignores our contemporary understanding of the communication process. In contrast, an alternative assumption is virtually all forms of communication are subject to being miscomprehended. Stated somewhat differently, there is not necessarily any correspondence between a given communication as conveyed by the source and that same communication as interpreted and remembered by the receiver. This is because the meaning a receiver extracts from a given communication consists of both directly asserted meaning (as expressed in the message) and meaning inferred by that individual receiver. These latter meanings are a unique function of each receiver's total sum of prior experiences and the set of expectations he brings to the situation (cf. Harris and Monaco 1978).
If the latter assumption is correct, then the fundamental question to be resolved is not "Does advertising cause miscomprehension?" but, rather "Is there any greater level of miscomprehension associated with advertising than is associated with other comparable forms of ms radii communication?" In other words, it is conceivable that miscomprehension of advertising may not be uniquely attributable to advertising per se, but may reflect a natural error rate common to all comparable forms of one-way mass communication.
Given the lack of hard benchmark evidence, the Educational Foundation of the American Association of Advertising Agencies decided to fund three investigations on the questions surrounding miscomprehension. This report represents a highly condensed summary of the basic findings generated by the first of these investigations. The primary objectives of the present investigation were:
1. To determine whether viewers do, in fact, miscomprehend televised communication. [Given that the major share of national advertising is devoted to television and regulator interest has been most evident in regard to this medium, attention was confined to televised communications.]
2. Assuming that some degree of miscomprehension is detected, to determine whether there is a "normative range" of miscomprehension associated with televised communications.
3. To determine whether commercial advertising tends to have a rate of miscomprehension which differs from other forms of televised communication.
4. To determine whether viewers possessing certain demographic characteristics are more prone to miscomprehending televised communications than are others.
Given these objectives, three core concepts needed to be operationalized. These were: (1) communications, (2) viewers, and (3) miscomprehension.
Sixty different communications which had been broadcast over TV during the six months spanning the last quarter of 1978 and first quarter of 1979 constituted the set of test communications. These communications were broadly representative of most material being broadcast over commercial television. The set of sixty communications can be grouped into these categories: commercial advertising (n=25), noncommercial advertising (n=13), and program excerpts (n=22). The commercial ads consisted of 18 ads for brands and services and 4 "image" or "corporate" ads. The non commercial ad group included 8 public service announcements and 5 ads for social causes. Finally, the group of program excerpts consisted of 18 excerpts from popular TV programs (including national and local news) and 4 excerpts from speeches and editorials.
The sample consisted of 2,700 people, 13 years old and older. Sampling quotas insured that there were equal numbers of males and females tested at each of 12 testing sites and that there were 386 respondents subsumed under each of the following age brackets: 13-17; 18-24; 25-34; 35-44; 45-54; 55-64; and 65+. Compared to U.S. Census Bureau data, the sample was found to be broadly representative in terms of years of formal schooling completed, household income claimed for the preceding year, marital status, and race. As a whole, the sample had a slight upward bias in terms of both education and income. That is, it claimed to be more highly educated and reported a higher level of prior year's income than did the population at large.
A six item miscomprehension quiz (patterned after Preston and Scharbach 1971) was administered after exposure to each communication. Each quiz consisted of six brief statements which sampled the universe of important information content regarding the object of that communication. Viewers were asked to indicate whether each statement was "True" or "False" based upon what was stated or implied in the communication. Two of the six statements on each quiz were accurate (i.e., "True"); four were inaccurate (i.e., "False"). Half related to objective facts (i.e., they were accurate statements of what had been directly asserted in the communication) and half to inferences which could be drawn from what was stated in the communication regarding the product (or product equivalent).
The procedure involved intercepting potential volunteers in the twelve shopping malls and inviting them to participate in the investigation. If the individual agreed and conformed to pre-established sampling quotas, s/he was invited into the testing room. Shortly thereafter s/he viewed the first of the two communications to which s/he had been randomly assigned and then responded to the miscomprehension quiz and a questionnaire regarding that communication. This was followed by viewing the second test communication assigned and responding to the set of corresponding questionnaires.
Several points should be made regarding the assessment procedures. First, interest focussed specifically and exclusively on miscomprehension and not on any of the prior or subsequent effects usually theorized to result from communication. Thus, the forced exposure procedure --to insure that exposure, awareness, and attention had actually occurred -- was considered appropriate.
Second, miscomprehension was assessed immediately after exposure to each communication (rather than after exposure to both communications) because of the rapid decay of memory traces found with such communications. As one example, Bogart (1967 p. 109-110) cites research to indicate that fewer than 201 of a group of 5275 respondents could recall the identity of the TV commercial they had viewed just a few minutes earlier.
Third, the procedures involved two sets of counterbalancing. On the one hand, the order of presentation of test stimuli was counterbalanced for main effects due to order. Each communication was presented a total of 90 time --first in 45 cases and last in the other 45 cases. On the other hand, so as to eliminate geographic biases, each of the 60 communications was tested at each site. Since each communication was involved in 90 separate tests, no less than 7 nor more than 8 of these tests conk place at any one of the 12 testing sites.
A series of multiple regression analyses were first conducted to determine whether variations in communication complexity, [Complexity was assessed via a battery of eight separate indices consisting of a copy point index, measurements of both visual and auditory complexity (see Watt and Krull 1974), and an assessment of the difficulty level of the verbal copy (using the Farr, Jenkins, and Paterson 1951, version of the Flesch count).] communication familiarity, product awareness-trial-usage, order of testing, or testing site exerted any appreciable effect on the miscomprehension scores obtained. In none of these analyses did these covariates cumulatively explain more than 8.4% of the variation in the scores. This suggests that miscomprehension is unaffected by variations in any of these factors.
Objective 1: Does Miscomprehension Occur?
The question "Does miscomprehension occur?" can be rephrased as three separate questions: (a) What proportion of viewers miscomprehend at least some portion of the communications which they viewed? (b) What proportion of the meaning in the test communications were miscomprehended? (c) What proportion of the set of 60 test communications were miscomprehended?
Proportion of Viewers who Miscomprehended. Only 16.8% of the viewers "fully comprehended" (i.e., were able to correctly answer all six quiz items for) either the first or second communication which they viewed. This means that in 83.2% of the 5400 viewings (2,700 viewers X 2 test communications each), viewers miscomprehended at least some portion of the communications which they viewed.
Correctly comprehending the content of one communication did not mean that that person was necessarily able to correctly comprehend another communication. In fact, of the 16.8% who fully comprehended one of the test communications (in the sense of answering all six quiz items correctly); most (13.31) miscomprehended some portion of the second communication which they saw. That is, only 3.51 of the viewers fully comprehended both test communications to which they were exposed (i.e., answered all 12 items correctly).
Proportions of Meanings Miscomprehended. One may also ask: To what extent was the content of the 60 test communications miscomprehended? The findings reveal that 29.6% of the (2700 respondents x 2 communications per respondent x 6 quiz item per communication =) 32,400 meanings that were tested were answered incorrectly. Both the median and model rates of Overall Miscomprehension were 28%. These figures probably underestimate the actual amount of miscomprehension, since they do not adjust for instances where viewers who did not know the correct answer managed to guess correctly.
Proportion of Communications Miscomprehended. It is noteworthy that at least some degree of miscomprehension was associated with each and every one of the 60 test communications. Across the set of 60 test communications, the level of miscomprehension ranged from a low of 11% to a high of 50%.
In sum, regardless of whether considered in terms of viewers, meanings, or communications, the answer to the first research question is a resounding "yes." Miscomprehension of televised communications does occur and seems to be far more prevalent than might have been anticipated. Perfect comprehension rarely occurs and may not be generally attainable.
Objective 2: What is the Typical Range of Miscomprehension?
The interquartile range (that is, the range encompassed by the middlemost 50% of the teat communications) was used to estimate the typical range of miscomprehension. [The interquartile range was employed in preference to "plus-or-minus one standard deviation from the mean" for two main reasons. First, the distribution of miscomprehension score was decidedly not normal. Second, the inter-quartile range lends itself more easily to layman comprehension.] Considering the entire set of 60 test communications, the inter-quartile range extended from 23% to 36%. Generalizing from these data suggests that one might expect anywhere from one-quarter to one-third of the material information content contained in communications that are broadcast over commercial television to be miscomprehended.
Objective 3: Is Advertising More Miscomprehended than Non-Advertising?
Program excerpts vs. commercial advertising vs. noncommercial advertising. Based on the sample of 60 test communications, the 22 program excerpts were associated with a significantly [All statements of "significance" refer to being statistically significant at the P=.01 level or better. These data have been independently checked and confirmed by our statistical consultant, Professor Jacob Cohen.] higher rate of miscomprehension (X=32.2%) than were either the 25 commercial (X=28.3%) or non-commercial (X=27.6%) advertisements. However, this difference is quite small and, though statistically significant, may be considered immaterial in practical terms.
Product-service advertising vs. entertainment information program excerpts. The broad three-category classification system has the potential to obscure what is perhaps the most interesting comparison of all, namely, the contrast between product-service advertising on the one hand, and entertainment-information program excerpts on the other. These are the two most frequently broadcast kinds of communication appearing on network television. The investigation examined 22 product-service advertisements and 18 entertainment and information program excerpts. Are these two kinds of televised communication associated with different levels of miscomprehension?
The answer is "yes." Entertainment-information program excerpts were associated with a small, but significantly higher rate of miscomprehension than were product-service advertisements. The mean levels of miscomprehension were 31.0% and 28.8%, and the interquartile ranges extended from 27% to 37% and 22% to 34%, respectively.
Objective 4: Are There Demographic Differences Associated with Miscomprehension ?
A number of socio-demographic characteristics were assessed. These included the respondents' sex, age, marital status, employment status, race, years of formal education, and household income for the preceding year. Only two of these factors -- age and education--were significantly related to miscomprehension. Both younger and older viewers were more likely to miscomprehend the communications, and miscomprehension appears to decrease (although not appreciably) as amount of formal education increases. However, while statistically significant, for all intents and purposes, theme relationships are so weak as to have little practical applicability.
Based upon the findings, six specific conclusions appear justified. First, a large proportion of the American television viewing audience tends to miscomprehend communications broadcast over commercial television. The vast majority (96.3%) of the 2700 respondents in this investigation exhibited some degree of miscomprehension. Second, it would appear that no communication is immune from being miscomprehended. Each and every one of the 60 test communications was miscomprehended at least some of the time by moue of the viewers.
Third, the average amount of miscomprehension associated with each of the 60 test communications was 30% (actually 29.61%). In other words, approximately 30% of the relevant informational content contained within each communication was miscomprehended. Fourth, as a preliminary estimate, the typical range of miscomprehension is 29.5% + 6.5%. That is, typical TV communications seem to have their content miscomprehended at anywhere from 23% to 36%.
Fifth, non-advertising communications were associated with higher levels of miscomprehension than were advertising communications. Of particular interest, excerpts of TV programs were miscomprehended at higher levels than were commercial advertisements (for products-brands-services). However, though statistically significant, these differences are practically trivial. Finally, for all practical purposes, no major demographic variables appear to be meaningfully associated with miscomprehension. Miscomprehension seem to be widespread throughout the populace, occurring at all age, income, and education levels in our society and to the same degree.
Selected Implications of Our Findings
Aeons other things, this investigation reveals that, once again, the mere provision of information does not automatically translate into this information having any effect, much less the intended effect (cf. Jacoby 1974; Jacoby, Chestnut, and Silberman 1977). Rather, the receiver/viewer brings a storehouse of past experience and an ongoing mental set to each communication transaction and tends to interpret, and misinterpret, communications in terms of these mental phenomenon. Given that it is not possible to eradicate either the influence of past experience or the individual's current mental set, it may well be impossible to eradicate miscomprehension. This leads directly to a set of implications for public policy makers, educators, and researchers alike.
Perhaps the user fundamental implication for policy makers is that, just became there is a demonstrable degree of miscomprehension associated with a particular advertisement does not necessarily mean that that particular advertising contains something out of the ordinary to provoke said miscomprehension. Such an assumption may be totally unwarranted, Indeed, the findings of the present investigation revealed that the rates of miscomprehension associated with advertising were significantly lower than the rates associated with typical program content. In other words, a certain proportion of miscomprehension may simply reflect a natural error rate associated with all types of televised communication. The ramifications that this implication has for such regulatory actions as cease and desist orders, corrective advertising, and affirmative disclosure orders are substantial.
Of course, a reasonable argument is that the consequences of miscomprehending most TV programs tend to be trivial when compared to the consequences of miscomprehending, and then acting upon, advertising. Hence, the rate for advertising should be lower than that for other forms of televised communication. While the authors might be persuaded that advertising miscomprehension rates are probably unacceptably high, they would find completely unreasonable any attempt to use "zero-based miscomprehension" as the criterion for evaluating advertising and as a basis for formulating public policy.
Perhaps the data from the present investigation might serve a useful function in this regard. Regulatory interest in correcting misleading and/or deceptive advertising might use the boundary separating the third and fourth quartile of communications as a triggering mechanism, so that advertising which exceeds this level, while not automatically considered deceptive or misleading, would be targeted for further scrutiny. Of course, any attempt to employ such a cutoff level would have to accommodate for the fact that the data adduced here probably underestimate the extent of miscomprehension present under normal viewing conditions. Further, it needs to be recognized that there may be instances (as in the case of over-the-counter and prescription pharmaceutical products) where miscomprehension rates of 102 or even 5% might be deemed totally unacceptable (see Jacoby and Small 1975).
There are parallel implications for advertisers. Given recognition of the fact that some degree of miscomprehension is likely associated with any ad and may, if detected at sufficiently high levels, become the focus of regulatory attention, advertisers might be encouraged to pressure their advertising departments and advertising agencies for lower rates of miscomprehension. Aside from concern over possible regulatory activity, there is a very practical reason for engaging in such research. Namely, by lowering the rate of miscomprehension, the advertiser would insure that more of his advertising message is being accurately comprehended. Perhaps the syndicated advertising evaluation services could be encouraged to devise and provide a "miscomprehension index" along with the other standard communication impact indices which they currently do provide.
Another implication -- one which is consistent with much previous research -- is that broadcast advertising may not be a suitable medium for communicating substantial amounts of product information, particularly complex product information. In terms of the classical hierarchy-of-effects conceptualization (Lavidge and Steiner 1961) and more contemporary models of consumer decision making (e.g., Engel, Blackwell and Kollat 1978), perhaps the basic function of advertising is simply to stimulate awareness and generate problem recognition. Labeling and other information media would seem to be better vehicles for providing the detailed supporting information
Some will no doubt quarrel with the procedural details of this investigation or with the specific findings obtained arguing, as examples, that the 302 miscomprehension rate is too high to be correct, or that commercial advertising is more readily miscomprehended than is TV programming. It may well be that these arguments are correct. No claim is made of providing definitive answers to the research questions which stimulated this investigation. The authors merely claim that the findings represent preliminary ballpark estimates which are in need of further confirmation and refinement. However, the fact remains that, even with a procedure -- involving unclutter forced exposure and volunteer participants who were likely more attentive to the test communications than they normally would otherwise have been -- which should have reduced the rate of miscomprehension, nearly every single one of the 2700 respondents managed to miscomprehend at least some portion of the communications which they viewed, each and every one of the 60 test communications was miscomprehended, and the average amount of miscomprehension associated with any given communication was an unexpectedly high 30%. Further, these findings were quite robust, holding true for respondents of different ages, incomes, education levels, sexes, and marital status.
The implications of these findings are substantial. Perhaps foremost among these is that to require advertising -- or any other comparable form of communication, for that matter -- to conform to a "zero base" miscomprehension rate appears totally unreasonable. Further, to contend that advertising is necessarily more prone to being miscomprehended than are other forms of televised communication is similarly unreasonable. If nothing else, the present findings strongly suggest the need for a programmatic series of research investigations designed to shed considerably greater light on the various questions surrounding viewer miscomprehension of televised communications.
Bogart, L. (1967), Strategy in Advertising. New York: Harcourt Brace, Chapter 5.
Engel, J. F., Blackwell, R. D., & Kollat, D. T. (1978), Consumer Behavior (3rd ed.). Hinsdale, Illinois: Dryden Press.
Farr, J. N., Jenkins, J. J., & Paterson, D. G. (1951), "Simplification of Flesch Reading Ease Formula." Journal of Applied Psychology, 35(5), 333-337.
Harris, R. J. and Monaco, G. E. (1978), "Psychology of Pragmatic Implication: Information Processing Between the Lines." Journal of Experimental Psychology: General, 107(1), 1-22.
Jacoby, J. (1974), "Consumer Reaction to Information Displays: Packaging and Advertising." In S. R. Divita (Ed.), Advertising and the Public Interest. Chicago: American Marketing Association, pp. 101-118.
Jacoby, J., Chestnut, R. W., and Silberman, W. (1977), "Consumer Use and Comprehension of Nutrition Information." Journal of Consumer Research, 4(2), 119-128.
Jacoby, J., Hoyer, W. D., & Sheluga, D. A. (1980), The Miscomprehension of Televised Communication. N. Y. American Association of Advertising Agencies.
Jacoby, J. & Small, C. B. (1975), "The FDA Approach to Defining Misleading Advertising." Journal of Marketing, 39(4), 65-68.
Lavidge, R. J. & Steiner, G. A. (1961), "A Model for Predictive Measurements of Advertising Effectiveness." Journal of Marketing, 25(4), 59-62.
Preston, I. L. & Scharbach, S. E. (1971), "Advertising: More Than Meets the Eye?" Journal of Advertising Research, 11(3), 19-24.
Watt, J. & Krull, R. (1974), "An Information Theory Measure for Television Programming." Communication Research, 1(1), 44-68.