Source Effects, Message Effects, and General Effects in Counteradvertising


H. Keith Hunt (1972) ,"Source Effects, Message Effects, and General Effects in Counteradvertising", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 370-381.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 370-381


H. Keith Hunt, The University of Iowa

In January of 1972, the Federal Trade Commission (FTC) in a brief to the Federal Communications Commission (FCC) supported the concept of counteradvertising as a suitable approach to remedying some of the present failings of advertising. Counteradvertising is advertising which presents an opposing viewpoint to an ad that has already been run which the counteradvertiser finds objectionable in some way. This suggestion of the appropriateness of counteradvertising is the most recent in the series of actions taken by the FTC to inhibit the incidence of and offset the effects of false, misleading, and/or deceptive advertising on consumers and on competition in general. As with the previous introduction of corrective advertising, counteradvertising is proposed with no research evidence to support it.

Little is known about the effects of counteradvertising, even as to whether it has a positive or negative effect on attitude toward the brand. The FTC made its suggestion with no supporting evidence of the probable effects of the practice. The need for research into the effects of counteradvertising led to the following experiment designed to (1) investigate the effects of counterads, in general, on attitudes, (2) investigate the effects of different sources of counterads (consumer organizations, competitors, government agencies, and corrective ads by the original advertiser), and (3) investigate the effects of different message formats (unsupported statements as compared to supported, research-based statements). Findings from the investigation will give an indication both of the overall efficacy of counteradvertising and of which sources and message content choices are most effective.

The experiment exposed all subjects to the same advertisement and then introduced the various counterad treatments in the following 2 x 4 factorial design. Two message forms were used for the counterads: a supported statement and an unsupported statement. Four sources were used: a consumer organization, a competitor, a government agency, and the original advertiser. The original advertiser issued a "required corrective statement" which stated that the original ad was deceptive. The other sources offered the counter message under their name alone. Also, a control group was exposed to the allegedly deceptive ad only, providing a base for determining how much change the counterads caused. The experimental design is as follows:



It was expected, based on previous research on the effects of corrective advertising (Hunt, 1972), that the supported message would cause a significantly greater decrease in favorableness of attitude than would the unsupported message. Also, it was expected that the noncommerCial sources would be more effective than commercial sources because of greater perceived truthfulness, and greater perceived sincerity.



A total of 270 subjects participated in the experiment and completed usable questionnaires. These were all unpaid students (sophomores, juniors, and seniors) in the Introduction to Marketing and the Computer Applications in Business courses at The University of Iowa.

Experimental Materials

The experimental materials consisted of several advertisements which were print ads of magazine quality, printed specially for this study.

The Allegedly Deceptive Ad. The allegedly deceptive ad used for the study was the original ad used to introduce Chevron's additive, F-310. This particular ad includes, according to the FTC, both misleading statements and deceptive presentation. The allegedly deceptive ad announces the new additive, F-310, and states that the additive reduces dirty exhaust emissions and thus improves gasoline mileage and helps toward cleaner air. Two pictures show balloons attached to cars' exhaust systems. The balloon attached to the car using Chevron with F-310 is clear. The other balloon is blackish from dirty exhaust emissions. This is offered as proof that Chevron with F-310 cleans the engine and reduces pollutant emissions (after being used for 2000 miles). Previous research into corrective advertising based on this ad found the ad to be easily understandable and meaningful to subjects (Hunt, 1972). Since Chevron products are not marketed in the geographic area where this study was run, the ad and counteradvertisements based on it would not affect the market for Chevron products. The openness of Chevron in allowing its controversial ad to be used for experimentation is appreciated.

The Chevron F-310 ad was used because it is an actual example of an ad counteradvertising would have been used to offset. Several groups were actively opposed to the ad before the FTC finally stepped in to require the corrective ad. The corrective ad requirement is currently in the hearing stage. It is possible that, except for the charge of false presentation, the rest of the claims made in the Chevron ad will not be disproved. Given the charges against the company's advertising and the controversial nature of the ad, Chevron is to be commended for its openness and willingness to allow the ad to be used in experimental studies.

The Counterads. Two forms of counterad were used, one form being an unsupported statement and the other being a supported, research-based statement. The unsupported statement has a headline reading "F-310 CLAIMS BY CHEVRON ARE FALSE AND DECEPTIVE ACCORDING TO (source)". Then, in the center portion of the page in large print is the following general statement: "TEST RESULTS SHOW THAT CHEVRON WITH F-310 DOES NOT CLEAN ENGINES AND DOES NOT REDUCE POLLUTION FROM UNBURNED HYDROCARBONS AND CARBON MONOXIDE EMISSIONS." At the bottom of the page is, "A PUBLIC SERVICE ADVERTISEMENT TO CONSUMERS FROM (source)". Since no evidence is given to support the counterad statement and since the message was the same except for the source, any difference in effect would be source effect.

The supported research-based statement carried the same headline, "F-310 CLAIMS BY CHEVRON ARE FALSE AND DECEPTIVE ACCORDING TO (source)" but was followed by supposed research findings by independent research firms showing that Chevron with F-310 had no effect on reducing polluting exhaust emissions or cleaning the engine. These research findings were made up for this experiment and were not real. Four independent studies were referred to, each presenting evidence that Chevron with F-310 did not do what it claimed to do in its advertising. The closing statement was "THE EVIDENCE SHOWS THAT CHEVRON WITH F-310 DOES NOT REDUCE POLLUTION," followed by the same bottom line "A PUBLIC SERVICE ADVERTISEMENT TO CONSUMERS FROM (source)." Again, since the supported statements were the same except for the source, any differences in attitudes across sources would be source effects.

Since each source gave both a supported and an unsupported form of counterad, a comparison of results for each source would indicate the relative effectiveness of the supported as compared to the unsupported message format. Since the unsupported counterad had no evidence to support it, it was accepted primarily on the credibility of the source. The supported counterad had the same source effect plus the strength of the research findings from the four independent research firms. Thus, the supported counterad should be a stronger attack than the unsupported counterad, resulting in lower levels of attitude toward the deceptively advertised brand.


Favorableness of attitude toward the brand was measured on the following 29 point scale. The subject expressed a degree of agreement or disagreement with a particular statement by circling the one dot in the line of dots which best represented his attitude concerning the statement.



In each case, the higher the score on the measurement scale, the stronger the agreement with the statement. The attitude statements for which measures were obtained are:

1. I like Chevron with F-310.

2. I think claims made by Chevron with F-310 were truthful.

3. I think Chevron is an expert source of information on gasoline and gasoline additives.

4. I think Chevron is sincere in its advertising about Chevron with F-310.

5. I think the claims made by (source) were truthful.

6. I think (source) is an expert source of information on gasoline and gasoline additives.

7. I think (source) is sincere in its advertising about Chevron with F-310.

General Procedure

The experimenter was introduced to the class as a professor who was doing a study of what people thought about different kinds of advertisements. The professor explained that the study would take about 15 minutes and that it involved carefully reading one or two ads and answering a short questionnaire about them.

The materials were in a 9 x 12 envelope which was passed out in a manner to assure that each cell was equally represented in each class group. The advertisements were stapled together in one packet with a cover and spacer sheets. The measurement scales were in another. Subjects were asked to take out the ad packet and turn to the first ad, reading it carefully until told to turn to the second ad. This took five minutes. Then the subjects turned to the second ad and read it for five minutes. Control subjects had no second ad (counterad) and were asked to wait for the next materials. Then all the ad packets were put back in the envelope and the measurement questionnaire taken out and completed. Students were told to put their names on their questionnaires. The questionnaires were put back in the envelope and the envelopes collected. Then the professor thanked the participants and left the room.

Statistical Procedure

Analysis of variance was used to test for significance of main effects. Tests of significance between cells in the experimental design were computed using Tukey's HSD test (Kirk, 1968). In the discussion section which follows, "not significant" means not significant at the .05 level of significance. "Significant" will have the level of significance indicated.


The control group data indicate that exposure to the Chevron ad resulted in a slightly favorable attitude toward the brand. However, counteradvertising in all but one case led to a substantially and significantly less favorable attitude toward the brand. The only exception is the competitor's unsupported attack which is discussed later. The findings are as follows:



A presentation of the findings in graphic form is presented on the following page.

In answer to the first research question about the general effects of counteradvertising, counterads do substantially reduce favorableness of attitude toward the brand. In seven of the eight cells the attitude level was significantly (p < .01) less favorable after a counterad exposure than before. Only in cell 2,2 was the cell mean not significantly different from the pretreatment, control group mean.

To answer the second research question about source effects, different sources of counterads did not make any difference except for the specific case of competitor-unsupported ads. While the source main effect was significant at the .05 level, this significance was due to the extreme variation in cell 2,2. The means were almost identical in the other seven cells.

To answer the third research question about message effects, supported and unsupported statements have the same effect except, again, in the case of cell 2,2, the competitor-unsupported ad, which was not significantly different from the control mean. While the message content main effect was significant at the .10 level, this significance was due to the extreme variation in cell 2,2,. As is seen in Figure 4, the means for each source, except for cell 2,2, were almost identical.

Except for one of the eight treatments, neither source nor message content differences made any difference in liking for the brand--each combination resulted in the same slightly to moderately unfavorable level of attitude toward the brand.


Source Effect

The similarity of source effects is surprising. It was expected that the consumer organization and government agency, the noncommercial sources, would effect the greatest decrease in attitude, that the corrective ad would effect less of a decrease and that the competitor would effect the least decrease.

It seemed reasonable that the noncommercial sources would be perceived as having less personal interest in the matter and more of a public service orientation, and would also be perceived as more sincere and more truthful than the commercial sources. Only in the dimension of expertness were the commercial sources expected to have a favorable advantage. Based on these expectations the noncommercial sources were expected to effect a greater decrease in favorableness of attitude than the commercial sources. Yet the four resulted in essentially the same level of attitude.

A review of the measures of truthfulness, expertness, and sincerity, revealed that all the predictions were borne out. Noncommercial sources were perceived as more sincere and more truthful. After exposure to a counterad, non-commercial sources were even seen as slightly more truthful. But the end result was nearly identical attitude levels for all sources. Further investigation is needed into which dimension or causes were strongest and what interactions occurred to yield this unexpected outcome.



Effects of Exposure to Counteradvertising on Perceived Truthfulness, Expertness, and Sincerity

It is interesting to note the changes in perceived truthfulness, expertness, and sincerity after exposure to the counterad. Perceived truthfulness of the sources was about the same before and after exposure to the counterad except for the corrective ad situation. Being required to admit in one's own ad that previous ad claims had been false caused a substantial drop in the perceived truthfulness of the corrective advertiser. After exposure to the counterad, perceived expertness rose for the noncommercial sources and dropped for the commercial sources.

Perceived sincerity dropped substantially for all sources of counterads, except the competitor, who wasn't perceived as very sincere in the first place. The noncommercial sources and the brand itself each dropped about equally in the sincerity rating. Evidently sources pay a price of decreased perceived sincerity when they engage in counteradvertising efforts. While expected for the brand, this is an unexpected finding for the noncommercial sources.

Previous research on the effects of corrective advertising showed perceived sincerity of the FTC to be unaffected by requiring a company to run corrective advertising (Hunt, 1972). Overall attitude toward the FTC actually increased after requiring the corrective ads. The decreased sincerity of both noncommercial sources is worth further investigation, but stands as a warning for the time being that counteradvertising may have deleterious effects for the counteradvertiser as well as the original advertiser.

Message Effects

Previous research on corrective advertising established that the supported attack was much more effective than the unsupported attack in reducing favorableness of attitude. In both studies the supported statement has been a pointed, precise attack on the falseness of the Chevron advertising claims. Also, in both studies the unsupported statement has merely stated that previous advertising by Chevron was false and misleading but no evidence or other specifics were mentioned. In the corrective ad study overall attitude was very little different after the unsupported statement attack than when no attack occurred. But the supported attack resulted in a substantial decrease in attitude. In the current study a much lower initial favorableness occurs, with equal drops in attitude from both formats, but not as much of a drop as the supported previously obtained. That the experimental situation was somehow different is likely, but subjects were similar, and ad formats were very similar. Nevertheless, the similar effects of both message formats is unexpected. It doesn't seem to make any difference which message is used.

The Competitor-Unsupported Case

The substantial differences in attitude level evoked by the competitor's counterads, with supported being much more effective than unsupported, is of interest. The measure of economic interest shows the competitor was considered very differently than the noncommercial sources--being perceived as interested in counteradvertising for his own advantage and not especially for the public interest. The result is that the competitor is a noncredible source not effective in reducing favorableness of attitude very much. Yet the supported attack was very effective. This strong effect is attributed to the research findings and their sources which are explicitly discussed in the ad. The four research studies, each by a group independent of the advertiser, were evidently so highly believable that they overwhelmed the competitor source effect. It is supposed that if the research had been done by the competitor rather than by independent research firms, the statements would have been less effective in reducing attitude. In the unsupported attack condition, no external source existed and the acceptance of the message rested solely on the credibility of the competitor, and the result was very little effect.









This previous point raises the question of whether the exceptionally strong supported statement with its four external sources not only swamped the competitor source effect but all four source effects. That all four sources had such similar attitude levels certainly supports such a contention. However, for the corrective source and both noncommercial sources, the unsupported attack based only on the credibility of the counteradvertiser source resulted in attitude level equally as low as the supported attack. It seems to be more a case of agreement between sources, counterad and research, than swamping, except in the case of the competitor-supported, in which the competitor source, being weak, was evidently swamped by the research sources mentioned in the statement.


It was expected (1) that counteradvertising would cause a decrease in favorableness of attitude, (2) that supported counter statements would be more effective than unsupported counter statements, and (3) that different sources would have different effects. Only the first expectation was borne out.

Several points should be kept in mind when interpreting and generalizing the present results. The study works with attitudes rather than purchase behavior, with some imitation rather than real ads, with lengthy forced single intensive exposures rather than several short happenstance exposures, with an unfamiliar product rather than a product currently available and being purchased, and with one product rather than several product categories. Also, students are far from being a cross section of gasoline purchasers. However the corrective advertising study found student responses to be very similar to results obtained from a cross section of the buying public. Finally, as presently conceptualized, counteradvertising would occur only in broadcast media--radio and television. Print ads were used in the study because of familiarity with the Chevron F-310 campaign and to allow for possible comparisons with the corrective advertising studies. The print ads were also more orderly to administer. Any studies eliminating one or more of these limitations would be a contribution to understanding the effects of counteradvertising. Hopefully, the government agencies involved in counteradvertising will some day feel a commitment to backing similar studies with their expertise and funds, and, as a result, basing their proposals on research findings as much as Possible .


Hunt, H. Keith. Deception, Inoculation, Attack: Implications for Inoculation Theory, Public Policy, and Advertising Strategy. Unpublished Ph.D. dissertation, Northwestern University, June, 1972.

Kirk, Roger E. Experimental Design: Procedures for the Behavioral Sciences. Belmont, California: Brooks/Cole Publishing Company, 1968.



H. Keith Hunt, The University of Iowa


SV - Proceedings of the Third Annual Conference of the Association for Consumer Research | 1972

Share Proceeding

Featured papers

See More


N1. The Experiential Advantage in Eudaimonic Well-being – An Experimental Assessment

Aditya Gupta, University of Nebraska-Lincoln
James Gentry, University of Nebraska-Lincoln

Read More


The Impostor Syndrome from Luxury Consumption

Dafna Goor, Harvard Business School, USA
Nailya Ordabayeva, Boston College, USA
Anat Keinan, Harvard Business School, USA
Sandrine Crener, Harvard Business School, USA

Read More


The Viciousness and Caring of Sharing: Conflicts and Motivations of Online Shamers

Chen Pundak, Tel Aviv University, Israel
Yael Steinhart, Tel Aviv University, Israel
Jacob Goldenberg, IDC

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.