IMAGE ADVERTISEMENTS FOR ALCOHOL PRODUCTS: IS THEIR APPEAL ASSOCIATED WITH ADOLESCENTS' INTENTION TO CONSUME ALCOHOL?

Kelly, Kathleen J. ; Edwards, Ruth W.

ABSTRACT

Criticism has been directed toward alcohol advertising, particularly regarding the use of image (lifestyle) advertising, and its potential influence on teenage alcohol consumption. This study sought to determine if adolescents who drink, or intend to drink alcohol at some future time, find image advertisements for alcohol more appealing than product advertisements. The results indicated that image advertising was preferred to product advertising, particularly by younger adolescents. Evidence of an association between preference for image advertisements and intent to drink in the future was found. Policy implications of the findings are discussed.

The controversy over the impact of alcohol advertising on alcohol consumption and attitudes about drinkers and drinking continues to mount. These debates have led to actual bans on alcohol advertising in some countries (Addiction Research Foundation, 1981) as well as to proposed legislation banning alcohol advertisements in others (Canadian Radio-Television Commission, 1990). Many proposals designed to minimize the impact of alcohol advertising have been put forth by legislators in the United States, with the requirement to place warnings in all forms of alcohol advertising receiving the most attention. Though the outcry directed at alcohol advertisers usually is concerned with the persuasive capabilities of advertising, more focused criticisms include the alleged targeting of underage drinkers through the use of appealing image advertising. Image advertising focuses on the lifestyle of the user of the product rather than on the intrinsic value of the product itself (Snyder & DeBono, 1985). The image adverti sement, with varying degrees of subtlety, suggests that the depicted lifestyles can be attained by use of the product. Image advertisements rarely make any explicit mention of product quality. In contrast, product advertisements focus on claims about the merit of the product (e.g., the taste is delicious or product performance is excellent).

A common response from the alcohol industry to criticisms of their advertising is that the focus of alcohol advertising is to encourage existing drinkers to maintain their brand preference, or to switch brands, and that it is not intended to attract new customers (Waterson, 1989; Mosher & Wallack, 1981). Further, alcohol industry spokespersons point out that research has never proven a causative link between advertising and consumption. The industry advocates continued self-regulation through the Advertising Standards Authority and controls emanating from the Independent Broadcasting Authority (Circus, 1989). In addition, the alcohol industry believes that the proposed requirement of warnings in alcohol advertisements is an infringement of their First Amendment rights (Dunn, 1991).

The majority of studies investigating the impact of alcohol advertising on consumption are econometric. Generally, variables such as advertising expenditures are manipulated in regression models to determine the impact on sales (Duffy, 1981; Ogborne & Smart, 1980; Bourgeois & Barnes, 1979). The results of these studies have been inconclusive and equivocal (Aitken et al., 1988; Comstogk, 1976). One of the limitations in using regression analyses of sales as a function of other variables is that sales analyses cannot reflect consumption by minors--the very audience that critics claim to be most vulnerable to alcohol advertising--since minors are not included in alcohol sales figures. The same limitation holds true in empirical and quasi experiments. Since these studies rely on observed purchasing behavior or consumption, it is not ethically or legally possible to measure the impact of advertising on adolescent consumption in this way.

Although self-reports of attitudes and behaviors are often criticized as not verifiable and therefore potentially unreliable, this methodology is one of the few viable alternatives for studying the impact of alcohol advertising on underage youth. Studies using this methodology most commonly investigate associations between exposure to advertising and consumption. Atkin, Hocking, and Block (1984) found significant associations between exposure to alcohol advertising and self-reported consumption of liquor among teenagers. Further, they found that among those adolescents who had not yet begun drinking, those with heavy exposure to alcohol advertising were more likely than their counterparts with more limited exposure to indicate that they plan to drink in the future (59% versus 36%). In a survey of 1,200 12- to 22-year-olds, a moderately positive correlation was found between the amount of day-to-day exposure to beer, wine, and liquor ads and both excessive alcohol consumption and drinking in hazardous situati ons (Atkin, Neuendorf, & McDermott, 1983). This association may, of course, be due at least partly to selective attention, with those who drink, or intend to drink, seeking out and paying greater attention to alcohol advertisements than do youth for whom alcohol use has less salience. These individuals may attend to alcohol advertisements more closely since the ads reinforce their already positive attitudes about alcohol and alcohol-related behaviors. A recent study by Grube (1993), which considered the level of awareness of television alcohol advertising rather than just exposure to alcohol commercials, was designed to address this issue. In this study, it was assumed that simple exposure to alcohol advertising has the same, effect on all children, and that advertisements have significant effects on knowledge, beliefs, intentions, and behaviors only when they are attended to, processed, and remembered. Grube found that children who were more aware of alcohol advertising had increased knowledge of brands and slogans and held more positive beliefs about drinking than did children who were less aware of the advertising. The increase in positive beliefs also mediated an increase in the intention of the children to drink as adults. These effects were maintained even though differences in background variables were statistically controlled. Grube's findings appear to support the hypothesis that alcohol advertising is a causal factor in the predisposition of children toward drinking.

In addition to general criticisms of alcohol advertising and its impact on youth, there are specific criticisms of the use of image advertising. The theories underlying self-monitoring and imaginary audience ideation would suggest that adolescents might have a stronger preference for image advertisements than would adults, and that females might have a stronger preference for image advertisements in earlier rather than later adolescence (Linn, de Benedictis, & Delucchi, 1982). Research has shown that as children move into adolescence, they develop an appreciation of the symbolism and imagery in commercials (Aitken et al., 1988; Aitken, Leathar, & O'Hagan 1985). As noted above, advertisements for alcohol in general have been found to become increasingly salient and attractive to youths between the ages of 10 and 14 (Aitken et al., 1988). Further, studies have concluded that adolescents find image advertisements for alcohol to be much more appealing than product advertisements (Covell, 1992).

Although self-reported alcohol consumption by adolescents has been studied by many researchers over the years, investigations into the link between consumption of alcohol or intention to consume and a preference for image advertising have been lacking. The present study tested the hypothesis that there is a positive relationship between the preference for image advertising (versus product advertising) and intention to consume alcohol by adolescents. Nationally, findings from the American Drug and Alcohol Survey [TM] (RMBSI, Inc., 1994) have indicated that intention to drink is higher than actual drinking behavior in the 7th grade, but by the 9th grade drinking behavior increases significantly and continues to rise through late adolescence. Since there is such disparity in drinking behaviors across this age range, students in the 7th, 9th, and 11th grades were included in the study in order to get a cross-section of drinking behaviors and intention to drink.

METHOD

Subjects

A total of 1,058 adolescents from three communities in the northeastern United States, with populations between 30,000 and 60,000, participated in the study. Participants included 376 7th graders, 376 9th graders, and 315 11th graders, of which 52% were male and 48% were female. Data from a few students were not used due to incomplete responses.

Instruments and Advertisements

Alcohol use questions from the widely used American Drug and Alcohol Survey[TM] were utilized to assess alcohol consumption and intention to consume. The survey's measures have been developed, refined, and tested in more than a decade of major research studies with populations that include 6th graders to young adults. Questions are simply worded and easily understood. Level of alcohol involvement is assessed by an eight-item scale. Cronbach's alphas of 93 for White Americans, .91 for African Americans, and .92 for Mexican Americans have been found for 7th through 12th graders.

A paired-comparison procedure was used to test the appeal of image versus product advertising. A pool of more than fifty pairs of advertisements that clearly represent either product or image advertising was assembled by a graduate marketing student. This student was instructed to select only full-page, full-color print advertisements for products with little market saturation to avoid preexisting perceptions and preferences (for example, brand names such as Lindemans, Grolsch, and Scores by were chosen). Additionally, to minimize previous exposure, he was to select the advertisements from current magazines that do not target the youth market. Only visually appealing product ads were chosen to increase the probability that the image ads were being selected for their "imagery" and not simply because they had more visual appeal than the product ads. Two panels of teenagers were then assembled in focus groups. After initial instruction in what constituted a product ad versus an image ad, and discussion to make sure all participants were clear on these definitions, the ad pairs were presented. These teenagers were then instructed to select the advertisements that, in their opinions, best met the definitions of product and image advertising. From the pairs of advertisements chosen, the investigators picked six for use in this study, with each pair representing a single product category (e.g., one pair for beer, one pair for wine, one pair for scotch).

Ad preference was assessed by having participants complete a short questionnaire for each ad pair. First they were asked which ad they would pay more attention to if reading a magazine, with response options being "the one on the left," "the one on the right," "equal attention to both," and "I wouldn't pay attention to either one." Next, they were asked to indicate which ad they liked better and the strength of their preference. Strength of preference was assessed by asking "how much more" they liked the one they chose, with response categories of "a lot," "somewhat," and "not at all." The fourth question asked which of the two advertisements in the pair they thought would be more likely to persuade someone to try the product, and the fifth question had them indicate which ad they thought their friends would like better (Linn et al., 1982, found that adolescents expect others to be more easily influenced by ads). The final two questions asked which product they thought cost more and if they had ever heard of the brand names of the products in the ads.

Procedure

The study was conducted in junior and senior high schools during regular school hours. The alcohol use survey was administered to the adolescents first to avoid any possibility that viewing the advertising would influence these data. Students were assured that their responses would be completely anonymous. Further assurances of anonymity were provided by explaining that no one would be walking around the room, and if they had a question, they would need to approach one of the "guests" (researchers). When completed, each student put his or her own alcohol survey into a large box in the middle of the room to prevent anyone else from seeing the responses during collection. After all alcohol surveys had been turned in, the advertising component of the project was initiated.

Students were shown one pair of advertisements at a time, presented on 12-inch by 15-inch poster boards which they could handle, allowing for close review, as well as on slides projected on a screen at the front of the room. The slides helped to ensure that subjects were evaluating the correct pair of advertisements at all times. The order of presentation of the product and image advertisements was mixed, such that each type appeared on the left side of the poster board and accompanying slide half of the time. The seven-item questionnaire was completed for each pair of ads while they were being viewed. Alcohol and evaluation questionnaires were coded so they could be matched.

The purpose of the study was not explained to the students until all the data were gathered. After the questionnaires were collected, the students were thoroughly debriefed, including a discussion of the controversy over the target audiences of image and product advertising. To counter any potentially negative consequences of the study (for example, stimulating the desire to drink), the debriefing included an overview of the dangers of teenage drinking. Students were given the opportunity to ask questions, and in some cases there was time to ask for elaboration from the participants on the rationale behind their advertising preferences. Antidrinking and antidrinking-and-driving media (e.g., posters) were shown to the students and left for each classroom.

RESULTS

Summary of Reported Drinking Behaviors

Reported drinking behaviors were compared with national figures to determine whether the participants were representative of youth in general. The rates of heavier drinking were somewhat higher among 9th and 11th graders in the present sample than in the national sample (the American Drug and Alcohol Survey[TM]; RMBSI, Inc., 1994). This elevation may at least partially be due to the fact that this study was conducted in the northeastern part of the U.S., where alcohol use by older adolescents, particularly in metropolitan areas, is higher than it is in other regions (Edwards, 1994). Level of perceived harm was lower than national averages, particularly with respect to regular alcohol use. Since perceived harm is a predictor of consumption, this finding was not surprising given the higher incidence of drinking behaviors reported.

As with national data on adolescent alcohol consumption, there was a clear upward trend in consumption by grade among the study population, with significant increases from 7th to 9th and 9th to 11th grades. The modal age at which those 11th graders who had gotten drunk reported they first had done so was 14. Of those 7th graders who did not consider themselves to be drinkers, 27% indicated that they will drink in the future. Of the 9th and 11th graders who did not consider themselves drinkers, 20% and 7%, respectively, indicated that they will drink in the future.

Students in all three grades indicated that beer was their preferred alcoholic beverage, and the majority of those who drank indicated that they liked to drink just a glass or two, or enough "to feel it a little." The degree to which adolescents drank to "feel it a lot" or "to get drunk" rose significantly with age. Further, perceived harm from getting drunk decreased with age.

Preference for Image Versus Product Advertisements

Image advertisements were, on average, preferred by participants over product advertisements for all pairs presented. For each pair of advertisements, respondents ranked the image advertisement as the one to which they would pay more attention, the one they liked better, the one they considered to be more persuasive, and the one they believed their friends would prefer.

Perceived cost of the product did not appear to be related to advertisement type. Participants selected the brands in the product ads as the more expensive in equal proportion to the brands in the image ads. This finding provides some validation for the choice of ads included in the study, which focused on selecting only highly appealing product ads.

Responses to ad pairs 1, 3, 5, and 6 were the most internally consistent and represented the clearest depictions of image versus product ads. Since the purpose of this study was to relate intention to drink to preference for ad type, it was appropriate to limit further analyses to these four ad pairs. Responses to these four pairs were utilized to generate an advertising preference index.

Intention to drink in the future. The measure of intention to drink captured both the adolescents' attitude and subjective norm which, according to behavioral intention theory, are the two best predictors of a behavior (Fishbein & Ajzen, 1975). Adolescents who indicated that they "don't drink and never will" or that they have used alcohol in the past but "don't plan to drink any alcohol again" were placed in the no-intention group. Conversely, adolescents who indicated that they "have had alcohol and probably will again" or "don't drink alcohol now but may" were placed in the intention group.

Advertising preference index. Four of the questions in the advertising survey were dichotomous, forcing respondents to select either the image or product advertisement. However, two questions allowed respondents to indicate the strength of their preference. From these questions a preference index was developed, with scores ranging from 4 (representing the strongest possible preference for the product advertisements) to 24 (representing the strongest possible preference for the image advertisements).

In addition to encouraging brand switching, the advertising industry claims that advertising is generally targeted to reach markets favorably disposed to a product category. Given this, perhaps the most susceptible adolescents are those who are favorably disposed to alcohol but have not yet begun to drink and establish brand preferences. Thus, mean scores for advertisement preference were compared with participants' intention to drink. Results of an analysis of variance (ANOVA) of preference index scores, by intent to drink, grade, and gender, are presented in Table 2.

Overall, adolescents reacted more favorably to the image advertisements than to the product advertisements. Although ad preference by intent to drink did not reach significance, there was a tendency for adolescents who intended to drink to prefer image ads. There were strong gender differences, but grade differences did not reach statistical significance. The significance of gender led to further analyses of the data, which showed that both males and females with intentions to drink had a stronger preference for image advertisements, except among 11th-grade females. The strongest preference for image advertisements for both genders was found among 7th graders. Males in every grade showed a stronger liking for the image advertisements than did females, but this preference was particularly strong among the 7th graders.

ANOVAs on the remainder of the survey questions by intent to drink were not significant. The majority of adolescents found the image advertisements to be more attention demanding, more persuasive, and more appealing to their friends, yet there was no significant relationship between these measures and their intent to drink. There were, however, some gender differences in the responses to the question asking which ad would more likely demand their attention. In 7th and 9th grades, males indicated that they would pay more attention to image advertisements than would their female counterparts (with the exception of one pair of ads). By the 11th grade, however, gender differences were not as evident. Interestingly, the persuasion question (which ad would be more likely to persuade someone to try the product) generated the most similar responses by gender. This may be due to the fact that the reference group ("someone" as opposed to "your friends") could be construed as less gender-specific.

DISCUSSION

This study found evidence that 7th, 9th, and 11th graders prefer image advertisements for alcohol to product advertisements. At all grade levels, participants indicated that they liked image ads better, they thought their friends would like image ads better, and they found image ads to be more persuasive and attention demanding. These findings are consistent with previous studies that considered the appeal of image versus product advertising (Snyder & DeBono, 1985; Covell, 1992) and the appeal of alcohol ads (Aitken et al, 1988). An exception is the study by Kohn, Smart, and Ogborne (1984), which found image ads to be no more effective or attention demanding than product ads. However, their study was conducted with adult male drinkers. Their results, in combination with the findings from the present study, lead to some interesting developmental questions. Since in this study adolescents' preference for image advertising weakened as they matured, it is plausible that by the time adolescents reach adulthood th ere is no longer a significant difference in ad preference. More research is needed to determine whether there is a developmental pattern to ad preference. However, since image advertisements for alcohol are so appealing to youth, and this is not the market advertisers profess to want to reach, perhaps product advertising should play a more significant role in alcohol advertisers' promotional mix.

In the present study, image ads were more appealing to males and females at all grade levels, with males showing a stronger preference than females, particularly among 7th graders. This is in contrast to a study by Covell (1992), which showed that in a general evaluation of alcohol ads, females found image ads to be more appealing than did males. However, Covell's subjects ranged in age from 8 to 16; the present study included subjects 13 to 17 years of age. The stronger preference among the males here is likely due in part to the use of female models in some of the image advertisements. Perhaps the ads used by Covell did not utilize models or, possibly, the preadolescent males in that study did not find the models as appealing as did the adolescent males in this study.

Some evidence was found for an association between those who drink, or intend to drink, and preference for image or product ads. Those who do not yet drink, but think they may in the future, showed stronger image ad preference. Several policy issues are raised by these findings. One issue that has received a fair amount of attention is that of banning image advertising for alcohol. In theory, banning image advertising should not impact sales of alcoholic beverages, since advertising's purported intent is to stimulate brand preference and not new demand. All advertisers would be under the same restrictions and the challenge would be to develop appealing product advertising. However, in reality, the banning of image advertising would be a short-term, band-aid approach, as advertisers would seek to develop increasingly effective product advertising. Once this product advertising was shown to appeal to youth, demands would surface to eliminate its use. Additionally, the demarcation between image ads and product ads is not always clear, which would create enforcement problems. Most important, however, it is unlikely that the banning of image advertising would quickly impact alcohol consumption among youth.

There are many environmental factors that are of greater significance to youth drinking than alcohol advertising (e.g., peer and parental sanctions). As long as alcohol consumption is socially approved by an adolescent's peer group and societal messages reinforce its use, teen drinking will persist at high and often dangerous levels. Counteradvertising could play a more significant role in contributing to a social climate that deglamorizes consumption and popularizes abstinence. Research with adolescents has helped us understand what youth find appealing about alcohol commercials (Aitken et al., 1988). This knowledge of the cues and imagery that are most appealing to adolescents should be used in the development of antidrinking messages. For example, instead of messages urging moderation, public service announcements could feature fun-loving, attractive youth engaged in exciting alternatives to drinking. To date, counteradvertising has not been particularly effective, since most of the efforts have been prog rammatic or have targeted specific issues, such as drunk driving or fetal alcohol syndrome. To be effective, a long-term, high-quality television campaign should serve as the cornerstone for such efforts, with local media reinforcing national messages.

Level of perceived harm of alcohol and illicit drugs is highly correlated with use and serves as a protective factor against abuse (Johnston, O'Malley, & Bachman, 1992). A recent review of research focusing on the design of warning labels and their impact on the perception of risk showed mixed results (Hilton, 1993). However, all of the studies reviewed examined perceived risks of specific actions, such as drinking during pregnancy, or the risks were assessed among adults and not adolescents. Perhaps warnings in alcohol advertising would be a more effective mode of message dissemination than would warning labels on products, particularly for adolescents who have not yet begun to drink. Legislation requiring such warnings in alcohol advertising was proposed in the Sensible Advertising and Family Education Act (Colford, 1993).

There are many questions the current study does not answer. Obviously, the associations found do not show causality. Atkin et al. (1984) have found that those who drink pay more attention to alcohol ads. Does the use of advertising imagery contribute to this increased attention and consumption? The present study also does not establish the basis of adolescents' ad preferences. As discussed previously, care was taken to ensure that both types of advertisements used in the study were comparable in their visual appeal. However, it is possible that peripheral cues, such as source attractiveness, influenced the adolescents' choices as much as or more than extensive thinking about the message (see Petty & Cacioppo, 1981, for a discussion of the central and peripheral routes to persuasion). Finally, this study was limited to print media; it is possible that use of broadcast media would produce different findings. It is difficult, however, to find recent television commercials that are clear depictions of product ad vertising for alcohol. Additionally, the televised brands would be well-known and there would likely be preexisting preferences.

In conclusion, debates over policy and proposed legislation associated with alcohol advertising will continue. Alcohol advertisers will attempt to focus the debate on First Amendment rights, and public health advocates will take a more utilitarian approach. The question ultimately becomes one of priority. In the absence of empirical evidence proving causation, should policy changes be implemented to restrict alcohol advertising?

Ruth W. Edwards, Ph.D., Research Scientist, Tri-Ethnic Center for Drug Abuse Prevention, Department of Psychology, Colorado State University.

Reprint requests to Kathleen J. Kelly, Ph.D., Assistant Professor, Department of Marketing, College of Business, Colorado State University, Fort Collins, Colorado 80523.

REFERENCES

Addiction Research Foundation. (1981). Alcohol: Public education and social policy. Toronto: Addiction Research Foundation.

Aitkin, P. O., Eadie, D. R., Leathar, D. S., McNeil, E. J., & Scott, A. C. (1988). Television advertisements for alcoholic drinks do reinforce under-age drinking. British Journal of Addiction, 83, 1399-1419.

Aitkin, P. P., Leathar, D. S., & O'Hagan, F. J. (1985). Children's perceptions of advertising for cigarettes. Social Science and Medicine, 21(7), 785-797.

Atkin, C., Hocking, J., & Block, M. (1984). Teenage drinking: Does advertising make a difference? Journal of Communication, 34(Spring), 157-167.

Atkin, C. K., Neuendorf, K., & McDermott, S. (1983). The role of alcohol advertising in excessive and hazardous drinking. Journal of Drug Education, 12, 313-325.

Bourgeois, J. C., & Barnes, J. G. (1979). Does advertising increase alcohol consumption? Journal of Advertising Research, 19(4), 19-29.

Canadian Radio-Television Commission urged to ban alcohol ads. (1990, October 26). Globe & Mail, p. Ba.

Circus, P. (1989). Alcohol advertising--The rules. International Journal of Advertising, 8, 159-165.

Colford, S., (1993, April 5). Call for alcohol ad warnings has friend in White House. Advertising Age, p. 3.

Comstogk, G. (1976). Television and alcohol consumption and abuse. Santa Monica, CA: Rand Corporation.

Covell, K. (1992). The appeal of image advertisements: Age, gender, and product differences. Journal of Early Adolescence, 12(1), 46-60.

Duffy, M. (1981). The influence of prices, consumer incomes and advertising upon the demand for alcoholic drink in the United Kingdom: An econometric study. British Journal on Alcohol and Alcoholism, 16, 200-208.

Dunn, W. (1991, July). The role of beer in our society: Responsible consumption. Executive Speeches, pp. 12-18.

Edwards, R. W. (1994). Alcohol, tobacco and other drug use by youth in rural communities. In S. Blaser, J. Blaser, & K. Pantoja (Eds.), Perspectives on youth violence and substances use in small towns and rural areas. Oakbrook, IL: Central Regional Education Laboratory.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior. Reading, MA: Addison Wesley.

Grube, J. (1993). Alcohol portrayals and alcohol advertising on television. Alcohol Health and Research World, 17(1), 61-66.

Hilton, M. E. (1993). An overview of recent findings on alcoholic beverage warning labels. Journal of Public Policy and Marketing, 12(1), 1-9.

Johnston, L. D., O'Malley, P. M., & Bachman, J. G. (1992). Smoking, drinking, and illicit drug use among American secondary school students, college students, and young adults, 1975-1991. Rockville, MD: National Institute on Drug Abuse.

Kohn, P., Smart, R., & Ogborne, A. (1984). Effects of two kinds of alcohol advertising on subsequent consumption. Journal of Advertising, 13(1), 34-48.

Linn, M. C., de Benedictis, T., & Delucchi, K. (1982). Adolescent reasoning about advertisements: Preliminary investigations. Child Development, 51, 1599-1613.

Mosher, J. F., & Wallack, L. M. (1981). Government regulation of alcohol advertising: Protecting industry profits versus promoting the public health. Public Health Policy, 2, 333-353.

Ogborne, A., & Smart, R. G. (1980). Will restrictions on alcohol advertising reduce alcohol consumption? British Journal of Addiction, 75, 293-296.

Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and contemporary approaches. Dubuque, IA: Wm. C. Brown.

RMBSI, Inc. (1994). The American Drug and Alcohol Survey [TM]. Fort Collins, CO: RMBSI, Inc.

Snyder, M., & DeBono, K. G. (1985). Appeals to image and claims about quality: Understanding the psychology of advertising. Journal of Personality and Social Psychology, 49(3), 586-597.

Waterson, M. J. (1989). Advertising and alcohol: An analysis of the evidence relating to two major aspects of the debate. International Journal of Advertising, 8(2), 111.