Television

Adolescents are heavy users of television. Extrapolating from recent data obtained from a nationally representative survey, 11- to 13-year-olds watch 27.7 hours and 14- to 18-year-olds watch 20.2 hours of broadcast and taped television programming each week (Roberts, Foehr, Rideout, and Brodie, 1999a). As a result, they are immersed in drinking portrayals and alcohol product placements. A recent content analysis of primetime television from the 1998-1999 season, for example, indicates that 71 percent of all programming depicted alcohol use and 77 percent contained some reference to alcohol (Christensen, Henriksen, and Roberts, 2000). Among those programs most popular with teenagers, 53 percent portrayed alcohol use; 84 percent of TV-14-rated programming, 77 percent of TV-PG programming, and 38 percent of TV-G programming depicted alcohol use. More episodes portrayed drinking as an overall positive experience (40 percent) rather than a negative one (10 percent), although negative consequences were mentioned or portrayed in 23 percent of episodes. Underage drinking was relatively rare. Only 2 percent of regular characters under the age of 18 were depicted drinking alcohol. In another recent content analysis, however, characters between the ages of 13 to 18 were found to account for 7 percent of all alcohol incidents portrayed (Mathios, Avery, Bisogni, and Shanahan, 1998). When it occurs, youthful drinking or expressed desire to drink is often presented as a means of appearing to be adult and grownup (Grube, 1995). Other research suggests that drinkers tend to be regular characters, of high socioeconomic status, attractive, and glamorous (Mathios et al., 1998; Wallack, Grube, Madden, and Breed, 1990), although youthful drinkers are depicted in a less favorable light than older drinkers. Drinking is often treated as humorous and is associated with valued outcomes such as camaraderie (Hundley, 1995). Although common when considered at the program level, the prevalence of drinking characters is considerably below that for the U.S. population. Thus, in a recent analysis of primetime programming, only 11 percent of characters over the age of 34 were drinkers compared with 52 percent of similarly aged adults in the U.S. population (Long, O'Connor, Gerbner, and Concato, 2002). Only 14 percent of characters between ages 18 and 34 drank and only 2 percent of those under 18 drank, compared with 61 percent and 19 percent, respectively, for the U.S. population in these age groups.

Little research has addressed the potential effects of exposure to drinking on television on young people's drinking beliefs and behaviors. Generally speaking, correlational studies have found small, but statistically significant, relations between television viewing and alcohol-related beliefs and behaviors. Thus, Tucker (1985) found that high school boys who were heavier television viewers drank more than lighter viewers. Similarly, Neuendorf (1985) reported that television viewing was related to beliefs about drinking among 10- to 14-year-old adolescents: Heavier viewers were more likely than lighter viewers to agree that people who drink are happy and you have to drink to have fun at a sporting event.

More recently, in a prospective study of 1,533 ninth-grade students, it was found that television viewing was related to initiation of drinking over an 18-month period (Robinson, Chen, and Killen, 1998). Specifically, each 1-hour increase in television viewing at baseline was associated with a 9 percent increased risk of initiating drinking during the following 18 months (OR = 1.09), after controlling for age, gender, and other media use. Unexpectedly, however, each hour of watching taped programming and movies on video was associated with an 11 percent average decrease in drinking initiation. Moreover, drinkers and nondrinkers did not differ in weekly hours of television viewing at baseline, and television viewing was not associated with increases in consumption among those young people who were already drinkers at baseline. A final study investigated reported television viewing and scores on a risky behavior scale that included drinking for a sample of 14-to 16-year-old adolescents (Klein et al., 1993). Although significant positive relations were found between viewing and involvement in risky behaviors for specific genres (e.g., cartoons), the results were inconsistent across genres and no effect was found for overall TV viewing. Moreover, data relating specifically to drinking were not presented.

These correlational studies suffer from potentially serious conceptual and methodological problems. Conceptually, none of the studies directly measured exposure to televised drinking portrayals. Rather, they relied only on measures of overall television viewing. The problem with such measures is that children watching equal amounts of television may be differentially exposed to alcohol portrayals depending on their program preferences and attention levels. More importantly, all of these studies used correlational analyses that cannot provide evidence for the direction of the relationship between television viewing and drinking beliefs and behaviors. Some unconsidered third variable may influence both viewing and drinking. This interpretation cannot be entirely discounted, even for the single longitudinal study.

In addition to the correlational studies, the influence of televised portrayals of drinking on young people has been addressed in experimental studies (Kotch, Coulter, and Lipsitz, 1986; Rychtarik, Fairbank, Allen, Foy, and Drabman, 1983). In both of these studies, children who were shown videotaped segments from popular television series containing drinking scenes expressed more favorable attitudes and beliefs about drinking than did children exposed to similar segments without drinking.

Although these studies are suggestive, they are problematic. First, the effects were small and selective. In one case (Kotch et al., 1986) significant effects were found for boys but not girls, and then only for a few of the measures of alcohol beliefs that were obtained. Second, the possibility exists that the children may have perceived the drinking in the video as representing the experimenter's expectations regarding their task in the experimental situation. Thus, they may have been responding to what they believed the experimenter wanted them to do, rather than the actual drinking scenes. Third, the experimental situation in both cases is highly artificial, making it difficult to generalize the results to the real world. Self-selection, differential attention, and other factors that operate in the natural viewing situation are not present. A major concern is the fact that exposure to the drinking portrayals in these studies is brief. The experimental situation simply cannot provide a parallel to the real world where exposure occurs more or less regularly over relatively long periods of time. It is likely that the cumulative effects of such long-term exposure are far more important than any effects of short-term exposure.

In sum, the available evidence regarding the influence of televised alcohol portrayals on young people is inconclusive, at best. Further research using more sophisticated research designs and analytic techniques will be necessary to provide a more definitive answer to this question.

Film

Adolescents spend considerably less time viewing movies and movie videos than they do television. Extrapolating from recent national survey data, 11- to 13-year-olds spend an average of 6.2 hours per week and 14- to 18-year-olds spend an average of 4.7 hours per week watching movies (Roberts et al., 1999a). In terms of alcohol content in films, recent content analyses indicate that alcohol was shown or consumed in 93 percent of the 200 most popular movie rentals for 1996-97 (Roberts, Henriksen, and Christensen, 1999b). Underage use of alcohol occurred in only about 9 percent of these films. Alcohol and drinking were presented in an overwhelmingly positive light. Drinking was associated with wealth or luxury in 34 percent of films containing alcohol references, and pro-use statements or overt advocacy of use occurred in 20 percent of these films. Anti-use statements appeared in 9 percent of films with alcohol references; 6 percent contained statements on limits as to when, where, and how much alcohol should be consumed; and 14 percent depicted refusals to drink. Drinking in film is often associated with risky activities such as crime or violence (38 percent), driving (14 percent), and sexual activity (19 percent).

Portrayals of negative consequences of drinking are relatively rare. In all, 57 percent of films with alcohol references portrayed no consequences to the user. Similar findings have emerged from other content analyses. Thus, at least one lead character drank in 79 percent of the top money-making American films from 1985 to 1995 (Everett, Schnuth, and Tribble, 1998). Moreover, 96 percent of those films contained references supportive of alcohol use whereas only 37 percent contained references discouraging alcohol use. Surprisingly, an analysis of all G-rated English-language, animated feature films available on video cassette revealed that 47 percent (38 of 81) depicted alcohol or drinking (Thompson and Yokota, 2001). Of the 81 films, 13 contained scenes set in bars or nightclubs. In 15 of the 38 films containing alcohol, some consequences were depicted, but in most cases these consequences were minor (hiccupping, staggering, flushing). None of the films contained an overt health warning about alcohol use, and good or neutral characters accounted for the majority of drinking portrayals (67 percent).

Studies on the effects on youth of exposure to depictions of drinking in films are rare. In one study (Bahk, 2001), college students were exposed to one of two versions of A Star Is Born, one of which depicted negative consequences of drinking for the lead character (e.g., performing poorly at a concert, fighting, dying in a drinking-related crash) and the other with the negative consequences edited out, leaving primarily positive consequences. The results indicated that viewing the positive consequences version, relative to the negative consequences version, led to more favorable attitudes toward drinking and to stronger intentions to drink. The effects were strongest for attitudes toward drinking for tension reduction and amusement and intentions to drink for stress management.

In a similar study (Kulick and Rosenberg, 2001), college students were exposed to a series of eight film clips with or without depictions of spirits consumption. Participants were randomly assigned to one of three conditions: a negative portrayal condition, a positive portrayal condition, or a control condition. In the negative portrayal condition, five of the clips contained drinking scenes portraying undesirable outcomes. In the positive portrayal condition, six of the clips contained drinking scenes with desirable outcomes. In the control condition, none of the clips contained drinking scenes. After viewing all of the clips twice, the participants completed measures of intentions to drink spirits and alcohol expectancies. Results indicated that participants in the positive portrayal condition had significantly more positive alcohol expectancies compared with controls, although they did not differ significantly from those in the negative portrayals condition (p < 0.09). The negative portrayals condition did not differ significantly from controls. In terms of negative expectancies, participants in both the positive and negative portrayals conditions had higher scores than the controls, but did not differ from one another. The groups did not differ in intentions to drink spirits in the next week.

In sum, alcohol portrayals are common in films, even in those with ratings indicating they are intended for children and adolescents. These portrayals are typically positive or neutral and drinking is associated with desirable outcomes and characteristics. Few studies have investigated the effects of film portrayals of drinking on young adults, adolescents, and children. The findings from these studies are mixed. Although evidence from one study shows that such portrayals can have small effects on drinking attitudes and intentions, the results from a second study are ambiguous. In addition, as with experimental studies of alcohol portrayals on television, it is not clear how relevant these studies are to the real-world viewing situation because they address only the short-term effects of limited and brief exposures in an artificial setting. Experimental demand also remains an issue for these studies.

Music and Music Video

Music, either radio or recordings, is a popular form of entertainment for young people. Thus, 11- to 13-year-olds spend 11.2 hours per week and 14- to 18-year-olds spend 9.3 hours per week listening to music on radio, CD, or tape (Roberts, 1999a). Research on alcohol-related content in song lyrics is comparatively rare. A recent content analysis (Roberts et al., 1999b) examined 1,000 of the most popular songs in 1996-97 across five genres of music popular with youth. This study found that 17 percent of all the lyrics contained references to alcohol and that alcohol was mentioned more frequently in rap music (47 percent) than in other genres such as country-western (13 percent), top 40 (12 percent), alternative rock (10 percent), and heavy metal (3 percent). Overall, 22 percent of songs with alcohol mentions referred to beer or malt liquor, 34 percent to wine or champagne, 36 percent to hard liquor or mixed drinks, and 31 percent to generic terms such as “booze.” A common theme was getting intoxicated or high (24 percent), although drinking was also associated with wealth and luxury (24 percent), sexual activity (34 percent), and crime or violence (13 percent). Consequences of drinking were mentioned in only 9 percent of the songs with alcohol references, and anti-use messages occurred in only 3 percent. Product placements or brand-name mentions occurred in 30 percent of them and were especially common in rap music (48 percent).

An analysis of alcohol depictions in rap music (Herd, 1993) found the portrayal of alcohol use to convey elements of disinhibition, rebellion, identity, pleasure, sensuality, and personal power. Similar to the ambivalent attitudes toward alcohol use expressed in country-western music (Chalfant and Beckley, 1977; Connors and Alpher, 1989), rap music vacillated between pro-drinking and anti-drinking attitudes. This analysis further associated the commodification of rap music by the malt liquor industry with an increase in malt liquor portrayals in rap songs.

DuRant et al. (1997) analyzed 518 music videos from MTV, BET, CMT, and VH1 for portrayals of alcohol and tobacco use. They found that portrayals of substance use varied by network and music genre, with MTV having the highest percentage of videos that portrayed alcohol and tobacco use, and CMT with the lowest percentage of tobacco use in videos. In terms of music genre, rap music videos contained the highest percentage of depictions of alcohol use, whereas rhythm and blues videos showed the least alcohol use. Additionally, alcohol use was found in a higher proportion of music videos that had any sexual content than in videos that had no sexual content.

In one of the only studies to address the relationship between music video viewing and alcohol use, Robinson et al. (1998) examined the association between media exposure and self-reported alcohol use. They collected baseline and 18-month follow-up data on media usage (television watching, video watching, playing computer or video games, and watching music videos) and lifetime and 30-day alcohol use for a sample of ninth graders. They found a 31 percent increased risk of drinking initiation within the next 18 months for each 1-hour increase in watching music videos. Although longitudinal in nature, causality is an issue in this study and the possibility remains that both exposure to music videos and initiation to drinking are related to a third unmeasured predisposing factor.

Images in Alcohol Advertising

Content analysis suggests that many alcohol advertisements link drinking with valued personal attributes such as sociability, elegance, and physical attractiveness and with desirable outcomes such as success, relaxation, romance, and adventure (e.g., Atkin and Block, 1981; Postman, Nystrom, Strate, and Weingartner, 1988; Strickland, Finn, and Lambert, 1982). Moreover, young people find some alcohol advertising appealing and are attracted to it. In a recent study (Chen and Grube, 2001; Martin et al., 2002), 450 fifth through eleventh graders were asked to rate 20 beer and soft drink advertisements. Overall, soft drink advertisements were rated more favorably than beer advertisements. Nonetheless, the young people identified some beer advertisements as being among their favorites in the 20 (Figure 11-3). In fact, the most liked advertisement overall was for beer and featured the Budweiser^® lizards and ferret. More than 90 percent of the young people liked this advertisement. Other beer advertisements were also popular, including a Bud Light advertisement featuring a computer-animated mouse (76 percent) and a Budweiser advertisement featuring the Clydesdale horses and Dalmatian dogs (84 percent). In general, children and adolescents find alcohol advertising with celebrity endorsers, humor, animation, and popular music to be particularly appealing (Atkin and Block, 1983; Chen and Grube, 2001; Martin et al., 2002). Adolescent boys are especially attracted to alcohol advertisements depicting sports (Slater et al., 1997; Slater et al., 1996). In general, adolescents and adults find lifestyle or image-oriented alcohol advertising to be more appealing than alcohol advertisements that promote only product quality (Covell, Dion, and Dion, 1994; Kelly and Edwards, 1998; Kelly, Slater, and Karan, 2002). Image advertising is especially preferred among younger adolescents (e.g., seventh graders) and particularly by younger males (Kelly and Edwards, 1998). Lifestyle or image-oriented advertising also appears to result in more favorable attitudes toward alcohol brands and products among young people when compared with strictly product-oriented or informational advertising (Kelly et al., 2002).

FIGURE 11-3

Percentage of fifth to eleventh graders who strongly like alcohol and soft drink advertisements.

Does alcohol advertising increase alcohol consumption and problems among youth? This question is addressed here by reviewing the recent research on alcohol advertising and by critically considering the evidence about the effects that exposure to these advertisements may have on alcohol beliefs and attitudes and on the prevalence of drinking and drinking problems among young people.

Does Alcohol Advertising Affect Drinking or Drinking Problems?

Earlier reviews have concluded that the evidence for the effects of alcohol advertising on drinking beliefs and behaviors is limited at best (e.g., Atkin, 1995; Calfee and Scheraga, 1994; Fisher, 1993; Nelson, 2001). The available research on the effects of alcohol advertising can be grouped into three types of studies: (1) experimental or laboratory studies, (2) ecological studies, and (3) survey and other correlational studies. Each of these types of studies will be considered in turn.

Experimental Studies

Experimental studies investigate the effects that short-term exposure to alcohol advertising under controlled conditions has on drinking beliefs and behaviors. Typically, a group of experimental participants will be exposed to one or more alcohol advertisements embedded within a television program, within a series of “neutral” advertisements, or, in the case of print advertising, within a booklet or magazine. The drinking beliefs or behaviors of this experimental group are then compared to a control group that watches the same program, sees the same collection of advertisements, or reads the same booklet, but without the embedded alcohol advertisements. The results of earlier experimental studies have been mixed with some studies finding no effects (e.g., Kohn, Smart, and Ogborne, 1984; Sobell et al., 1986) and other studies finding small or short-term effects for some study participants (e.g., Kohn and Smart, 1987).

Apparently only a single recent study has been published that experimentally manipulated exposure to alcohol advertising (Lipsitz, Brake, Vincent, and Winters, 1993). This study was intended to investigate the effects of television beer advertising on alcohol expectancies among young people who were not yet regular drinkers. Groups of fifth and eighth graders were exposed to videotapes containing five beer commercials, the same five beer commercials plus two antidrinking public service advertisements, or five soft drink commercials. Results of a memory task indicated that the children paid attention to the advertisements and remembered seeing the beer and soft drink commercials. Despite the attention given to the advertisements, however, neither exposure to the beer advertisements alone nor to the beer advertisements in combination with the antidrinking PSAs affected scores on the alcohol expectancy scales.

The results of these experimental studies offer only very limited evidence that alcohol advertising promotes more favorable drinking beliefs or increases consumption. Laboratory studies of alcohol advertising effects, however, can be criticized on at least four grounds (cf. Atkin, 1995; Grube, 1993; Lastovicka, 1995; Thorson, 1995). First, although laboratory experimental studies can control for extraneous factors and allow for strong causal inferences, they often lack realism. In the typical study, respondents will be exposed to alcohol advertising in an artificial setting (e.g., school-room) that does not resemble the natural viewing situation. As a result, it is difficult to draw conclusions about the “real-world” effects of alcohol advertising on beliefs and behaviors based on these laboratory studies.

Second, it has been noted that advertisers target specific audiences with particular advertisements (Thorson, 1995). If the stimulus advertisements do not contain images, themes, or music that appeal to the participants in a specific study, then it is unlikely that any effects will be observed.

Third, laboratory experiments on the effects of alcohol advertising can only address the effects of short-term exposure to a limited number of advertisements. The relevance of such studies for understanding the cumulative effects of exposure to hundreds or thousands of alcohol advertisements over many years is questionable. Laboratory studies may be more appropriate for studying short-term disinhibitory or priming effects than ascertaining if advertising has long-term effects on beliefs or behaviors. That is, this research paradigm may be most relevant to understanding if exposure to alcohol advertising elicits immediate and short-term increases in consumption among those already favorably predisposed to drinking (e.g., Kohn and Smart, 1987).

Fourth, it may be naive to expect alcohol advertising experiments, as typically implemented, to produce significant effects on beliefs or behaviors. In particular, it seems unlikely that exposure to a handful of alcohol advertisements in a laboratory setting could produce a measurable effect against the high background rates of such advertising to which respondents are already exposed in their everyday lives.

Ecological Studies

Generally, ecological or econometric studies have focused on the relationship between alcohol industry advertising expenditures and aggregated (e.g., per capita) alcohol consumption, sales, or problems. A few studies have investigated the effects of alcohol advertising restrictions or bans.

Survey and Other Correlational Studies

For the most part, survey studies of alcohol advertising have focused on children and adolescents. In general, the survey studies have addressed a fundamentally different question from those addressed in the ecological studies. Rather than asking if alcohol advertising affects overall consumption among young people, these studies ask who might be affected and by what processes. These are questions that cannot be addressed with aggregated data and the types of analyses typically used in ecological studies. In addition, rather than relying on measures of potential exposure at the population level (e.g., advertising expenditures), survey studies have focused at the individual level and specifically on young people known to be more or less exposed to, attentive to, or attracted to alcohol advertising. Early survey studies found small, but significant, positive relationships between reported exposure to alcohol advertising and drinking beliefs and behaviors among young people (Aitken, Eadie, Leather, McNeill, and Scott, 1988; Atkin and Block, 1981; Atkin, Hocking, and Block, 1984; Atkin, Neuendorf, and McDermott, 1983). These effects were small, however, and some studies failed to find substantively meaningful relationships between alcohol advertising and drinking beliefs and behaviors among young people (e.g., Adlaf and Kohn, 1989; Strickland, 1982 Strickland, 1983).

More recently, research has focused more on attentional and affective processes that may mediate between exposure to alcohol advertising and drinking beliefs and behaviors. Specifically, it has been proposed that attention to and positive affect toward alcohol advertising and the characters and images it contains may be factors that are important in determining whether alcohol advertising influences drinking beliefs and behavior (e.g., Austin and Nach-Ferguson, 1995; Grube, 1995; Grube and Wallack, 1994; Thorson, 1995). In one study a small sample of high school students were exposed to videotaped television beer advertising with and without sports content (Slater et al., 1997). The advertisements were embedded in either a sports program or an entertainment program. It was found that girls responded more negatively to beer advertisements and counter-argued them more than boys did, particularly when they had sports content. Of most relevance here, non-Hispanic white adolescents who were more favorable toward the beer advertisements were also more likely to report current drinking and future intentions to drink. The effects, however, were small and were not replicated among Latino adolescents. Moreover, because of the cross-sectional nature of the study, it is not clear what the causal relationship is. In particular, it cannot be determined if a favorable orientation to alcohol advertisements predisposes young people to drinking or if being predisposed to drinking makes young people more favorable toward alcohol advertisements.

The relationship between evaluation of alcohol advertising and drinking was also investigated in a study of 7- to 12-year-old children (Austin and Nach-Ferguson, 1995). It was found that liking alcohol advertisements was positively related to experimentation with alcohol. This effect was relatively robust, accounting for about 11 percent of the variance in the drinking measure. In a similar study with at-risk preadolescents, it was found that identification with the characters in alcohol commercials predicted expectancies regarding the positive effects of drinking (Austin and Meili, 1994). In a similar study, with third, sixth, and ninth graders, both rated desirability of characters in alcohol advertisements and identification with those characters predicted positive alcohol expectancies and, indirectly, a risky behavior index that included drinking (Austin and Knaus, 2001). These patterns of results were replicated and extended using a sample of ninth and twelfth graders from central California (Pinkleton, Austin, and Fujioka, 2001). A “predrinking behavior” index was also constructed by having the respondents rate the extent to which they would like to have each of a series of clothing and toy items with alcohol-related logos. Perceived desirability, identification, and ratings of advertisement production and content quality were found to be related to alcohol expectancies, the predrinking index, and an alcohol-behavior index. Again, however, all of these studies were cross-sectional and rely on simple correlational and regression techniques, thus precluding causal interpretations of these relationships.

Another study used survey data obtained from fifth- and sixth-grade school children (Grube and Wallack, 1994; Grube, 1995). Awareness of alcohol advertising was ascertained by presenting them with a series of still photographs taken from television commercials for beer. In each case, all references to product or brand were blacked out. They were asked whether they had seen each advertisement and, if so, to identify the product being advertised, and, if they knew that, the brand of the product. The most important findings from this survey were that awareness of advertising was related to increased knowledge of beer brands and slogans and to more positive beliefs about drinking. Awareness of alcohol advertising also had a significant indirect effect on intention to drink as an adult that was mediated through positive beliefs. Importantly, these effects were maintained even though the reciprocal effects of beliefs and knowledge on awareness of advertising were controlled through the nonrecursive modeling. The findings were interpreted as suggesting that awareness of alcohol advertising may predispose young people to drink rather than the other way around.

A series of recent studies from New Zealand have reached similar conclusions. In one study of 10- to 17-year-olds (Wyllie, Zhang, and Casswell, 1998b), respondents were given a written description of and shown a still photograph taken from three television beer advertisements. They were then asked how often they had seen each advertisement and how much they liked each of them. Liking was significantly related to intention to drink as an adult. This effect accounted for a substantial proportion of the variance in these drinking intentions. The effect of liking on current drinking behaviors was more modest and nonsignificant (p < 0.06). Importantly, the reciprocal effects of drinking and drinking intentions on liking were not significant. Stronger results were reported in a recent study of 18- to 29- year-olds (Wyllie, Zhang, and Casswell, 1998a) using similar procedures. In this case, liking of alcohol advertisements was positively related to endorsement of positive statements about drinking and to current alcohol consumption. Most importantly, liking of the advertisements was related to increased numbers of self-reported drinking problems. The reciprocal effects of drinking and drinking beliefs on liking of alcohol advertisements were not significant in the model.

Another recent survey study (Connolly, Casswell, Zhang, and Silva, 1994) represents an advance over previous studies because it used a longitudinal design. Recall of alcohol advertisements at age 15 was positively related to beer drinking 3 years later at age 18 among young men. Unexpectedly, among young women, those who recalled more alcohol advertising at age 15 reported drinking less beer at age 18 than did those recalling fewer advertisements. Although this study is important because it reports the first published longitudinal data showing a significant relationship between alcohol advertising and later drinking, it is problematic for a number of reasons. Most importantly, the analyses did not control for drinking or predisposition to drink at the earlier waves of the study. It is possible that those young men who recalled more advertisements at age 15 were already drinkers at that time and that the differences observed 3 years later were simply a continuance of this pattern. The fact that recall of alcohol advertising was related to decreased drinking among the young women further obscures the interpretation of this study.

A final survey study (Casswell and Zhang, 1998) investigated the relations among liking of alcohol advertising at age 18 and beer drinking and drinking-related aggression at age 21. The study used data from later waves of the previously described study (Connolly et al., 1994). The results indicated that liking alcohol advertising at age 18 was related to higher levels of beer consumption at age 21 and to alcohol-related aggression at that same age. The effects of liking of advertising on aggression were mediated through effects on drinking. Importantly, these effects were obtained even though drinking at age 18 was included as a predictor in the analyses. These findings are thus most consistent with the interpretation that alcohol advertising predisposes young people to drink, rather than drinking predispositions influencing attention to and affect toward alcohol advertising.

Preliminary analyses from other more recent studies (summarized in Martin et al., 2002) have replicated and extended these findings. Thus, one of these studies found that young people with more positive affective responses to alcohol advertising held more favorable drinking expectancies, perceived greater social approval for drinking, believed drinking was more common among peers and adults, intended to drink more as adults, and drank with higher frequency and in greater quantities. Interestingly, the effects of affective response to alcohol advertising on drinking behavior and intentions appear to be largely mediated through expectancies and normative beliefs. That is, a major consequence of alcohol advertising may be to increase young people's beliefs about the likelihood of positive consequences of drinking and the normativeness of drinking. In a second of these recent studies, exposure to alcohol advertising was found to be related to increases in drinking over time. The converse, that drinking would affect exposure to alcohol advertising, was not supported.

In sum, survey studies generally find significant associations between reported exposure to, attention to, and recall of alcohol advertising, on the one hand, and drinking beliefs and behaviors, on the other. These relationships, however, tend to be modest for the most part. Moreover, a number of these studies have used small and nonrepresentative samples, raising questions about generalizability. In addition, because of the cross-sectional designs of most of these studies and the failure to control for previous drinking in one of the longitudinal studies, it is difficult to make statements about causality. The emerging studies (Martin et al., 2002) are suggestive of advertising effects and may provide a more definitive answer once more complete analyses become available.