An advertisement’s production value refers to stylistic features that stimulate cognitive engagement. Both the Activation Model of Information Exposure (AMIE) (Donohew, Lorch, & Palmgreen, 1998) and the Limited Capacity Model of Mediated Messages (LCMMM) (Lang, 2006) state that, among youth and young adults, certain advertising features lead to greater, involuntary attention while also prompting more extensive message processing, higher recall, and more positive attitudes toward both the ad and the advertised product (Palmgreen et al., 1995; Stephenson, 2003), unless the overall cognitive demand is too high (Lang, 2000).

Important stylistic features include the following. Animation is appealing due to its distinct visual style and its association with cartoon entertainment (Nash, Pine, & Messer, 2009). Intense images—that is, images that are graphic or disgusting—capture attention and promote recall among teens and young adults by eliciting an emotional response (Donohew et al., 1998; Niederdeppe, Davis, Farrelly, & Yarsevich, 2007). Sound saturation refers to background sounds, such as street noise, that increase an ad’s realism and therefore its believability (Morgan, Palmgreen, Stephenson, Hoyle, & Lorch, 2003). Loud and fast music increases an ad’s sensation value and distinguishes it from contiguous advertising or programming (Morgan et al., 2003).

Storytelling formats are associated with improved recall (Mandler & Johnson, 1980), but also depress the audience’s ability to recognize an ad’s persuasive intent and to counterargue (Slater & Rouner, 2002). According to research by Chen et al. (2005), youth prefer ads that tell a story over those that simply present music, animals, and people but without a story narrative. Second-half punch means having a surprising ending, which is rated by youth as having high sensation value, a quality they often seek out (Morgan et al., 2003; Niederdeppe et al., 2007). Transitions to different camera shots, or edits, engage viewers because of the novel visual information that each shot presents (Lang, Zhou, Schwartz, Bolls, & Potter, 2000; Southwell, 2005). A faster pace of edits, up to the point of creating cognitive overload, increases the viewer’s physiological arousal and memory activation, which increases liking for the ad (Donohew et al., 1998; Lang et al., 2000; Niederdeppe et al., 2007).

Certain characters portrayed in ads can grab attention, stimulate audience identification, and increase liking for both the ad and the product. Perceived similarity—when characters and viewers share important traits—improves ad liking (Chen et al., 2005) and enhances modeling effects (Martin & Kennedy, 1994). The characters’ perceived attractiveness is also important. In one study, females, particularly those with low self-esteem, reported greater trust in and liking for ads with attractive models (Martin & Kennedy, 1994). Research has also shown that youth associate the glamour of a celebrity endorser with the advertised product (Ross, Campbell, Wright, & Turk, 1984). When similar, attractive, or famous people are shown drinking alcohol, youth are more likely to believe that this is expected normative behavior and to model it within their own social networks (Austin & Meili, 1994; Bandura, 1986; Rimal & Real, 2005). Also note that the use of animated characters, animals, and anthropomorphized creatures leads to higher attention, positive emotional response, and greater liking for an ad, particularly among younger teens (Chen et al., 2005; Nash et al., 2009; Waiters et al., 2001).

Ads using the genres of magic, fantasy, violence, and humor are better liked by youth and create positive associations with the product. Younger youth positively associate magic and fantasy with entertainment and play (Lewis & Hill, 1998), whereas violence, which tends to appeal more to older youth, captures attention by eliciting an emotional response (Rajecki et al., 1994). Humorous ads are the most highly rated by youth of all ages (Aitken, Leather & Scott, 1988; Chen et al., 2005; Nash et al., 2009; Waiters et al., 2001).

Product appeals present rational arguments for purchasing a product based on attributes such as taste, cost, quality, and health effects. Such advertising is less likely to trigger impulse buying (Petty & Cacioppo, 1986; Petty, Fazio, & Brinol, 2009). Youth exposed to these ads rate them less favorably and report less desire to purchase the product (Chen et al., 2005; Waiters et al., 2001). Youth are also less likely to recall such ads due to the higher level of cognitive involvement they require (Petty & Cacioppo, 1986; Slater & Rouner, 2002). Interestingly, one study found that youth tend to disbelieve product appeals about alcohol (Aitken et al., 1988).

Reward appeals claim that a product will result in positive experiential outcomes. Studies show that youth perceive even indirectly communicated associations between positive rewards and advertised products (Jones & Donovan, 2001), and that they favorably rate these types of appeals (Chen et al., 2005; Lewis & Hill, 1998; Nash et al., 2009). In this context, it should be noted that neurodevelopmental research has established that adolescents are more likely than children or adults to make decisions based on emotions rather than cognitions when experiencing emotional arousal (Crone, Van Duijvenvoorde, & Peper, 2016).

Adolescents experience greater negative affect than children and adults and more often engage in risky behaviors such as alcohol use to cope with distress (Gould, Hussong, & Hersh, 2012; Whalen, Jamner, Henker, & Delfino, 2001). Many youth expect alcohol to induce relaxation and happiness (Fleming, Thorson, & Atkin, 2004), and having such expectations during adolescence is associated with both alcohol use and alcohol misuse later in life (Cable & Sacker, 2008; Patrick, Wray-Lake, Finlay, & Maggs, 2009).

Portrayals of social rewards such as friendship and high social positioning can also shape youth expectations of alcohol use. Teens often experience high levels of self-doubt and anxiety in social situations and can be preoccupied with peer acceptance (Pechmann, Levine, Loughlin, & Leslie, 2005). Unfortunately, many youth overestimate the importance and frequency of drinking among their peers and therefore feel greater normative pressure to drink (Borsari & Carey, 2003). In addition, they commonly expect drinking to diminish their social inhibitions (Fleming et al., 2004). With this, promises of positive emotional experience, peer acceptance, and friendship have the potential to increase youth acceptance of alcohol advertising messages (Pechmann et al., 2005).

Another youth-appealing reward is the achievement of individuality and personal success through brand identification (Buijzen & Valkenburg, 2008; Carr, 2002). During adolescence, youth seek to distinguish themselves from their parents and form a personal identity, an important step toward full adulthood (Erikson, 1950). During this time, youth actively search for identities to emulate (Jacobson, Atkins, & Hacker, 1983). Those who adopt a brand-related identity—for instance, by owning alcohol-branded merchandise—are more likely to drink and to transition to heavy drinking (Hurtz, Henriksen, Wang, Feighery, & Fortmann, 2007; McClure, Stoolmiller, Tanski, Worth, & Sargent, 2009). Depictions of drinkers as successful may further reinforce positive alcohol expectancies and encourage drinking (Jones & Donovan, 2001).

Sexually provocative images can draw attention and promote recall by eliciting an emotional response (Donohew et al., 1998). For teens, exposure to alcohol ads with romantic or sexual themes might also create expectations that drinking will lead to sex (Pechmann et al., 2005; Strasburger, 2012; Waiters et al., 2001). In this context, it is important to note that drinking has been shown to be strongly associated with risky sex among youth (Cooper, 2002).

Risk-related content refers to media depictions of risky activities such as adventure, physical performance, and activities requiring alertness and coordination. Such content is expected to appeal to young males in particular (Buijzen & Valkenburg, 2008; Nash et al., 2009), who are more inclined than females, adults, and children to make risky decisions (Gardner & Steinberg, 2005; Giedd, 2008). In line with the AMIE, the fast-paced, high-energy activities shown in alcohol ads are attention grabbing and can stimulate a physiological response associated with liking (Donohew et al., 1998). Also note that studies of adolescents’ neural responses to viewing risky behaviors have found decreased neural activity related to cognitive control and response inhibition (Sherman, Payton, Hernandez, Greenfield, & Dapretto, 2016).

Risk-related content also includes representations of inappropriate alcohol use, such as drinking at socially unaccepted times of the day, making excuses for drinking, and prolonged consumption (Rhoades & Jernigan, 2013). These appeals seem designed to counteract the demands of responsible decision making and to create an expectation among inexperienced, impressionable teens that drinking is an anytime, anywhere, carefree experience (Pechmann et al., 2005; Rhoades & Jernigan, 2013).

This body of work on youth-appealing advertising content is limited in a number of ways. First, the field has lacked a tool to measure youth-appealing content, thereby limiting our ability to quantify youth-focused marketing efforts and to study the relationship between such content and youth beliefs, attitudes, and behavior. Second, although previous content analyses have found youth-appealing content in alcohol advertisements (Fielder, Donovan, & Ouschan, 2009; Finn & Strickland, 1982; Gunter, Hansen, & Touri, 2008; Pettigrew, Johnson, & Daube, 2012a; Pinsky & Silva, 1999), including in media with large youth audiences (Pettigrew et al., 2012b; Rhoades & Jernigan, 2013; Sloane, Wilson, & Gunasekara, 2012; Winpenny et al., 2012), much of this research was done during the previous decade or in other countries where the marketing and drinking culture are often very different from the United States today. Finally, past research has not been conducted at the brand level, making it difficult to test the relationship between exposure to specific types of brand appeals and specific brand preferences among youth.

To address these gaps, we compiled the content features described in the preceding into an index for measuring advertising content appealing to youth and then sought to address the following questions:

RQ1: To what extent are youth appealing features used in alcohol advertisements?

RQ2: To what extent are there content differences by brand popularity and alcohol type?

We analyzed a selection of alcohol advertisements aired on television shows popular among U.S. youth ages 12–20 years. Using data on the alcohol brand preferences of underage youth (Siegel et al., 2013), we tested whether the ads for more popular brands or certain alcohol types presented a greater amount of youth-appealing content. We expected to find that the majority of ads in this sample would have youth-appealing features, and that the more popular brands among underage youth would have the greatest amount of such content. Because past research has shown that youth prefer beer to other alcohol types (Johnston et al., 2012), we also expected beer ads to present more youth-appealing content than wine or liquor ads.

We used Nielsen Company (New York, NY)¹ data to identify 20 regularly scheduled entertainment television programs that had the largest youth audiences (ages 12–20 years) in 2010–2011. In total, 192 alcohol ads aired on these 20 programs during the time the Siegel et al. youth respondents were reporting the alcohol brands they consumed (December 2011—May 2012). We categorized these ads into two groups: (a) ads for brands found to be popular among youth (n = 90), defined as the 25 brands most consumed by youth, and which make up 50% of the youth market share (Siegel et al., 2013), and (b) ads for brands that were less popular (n = 102). We randomly selected 50% of the popular brands’ ads (n = 45) and 50% of the less popular brands’ ads (n = 51), totaling 96 ads.

Two researchers worked independently to code a 10% subsample of the ads (n = 9), after which they discussed their discrepant coding until they reached agreement and then refined the codebook. The researchers then coded a second subsample (15% of the remaining sample, n = 13). Intercoder agreement on the subsamples was high (Cohen’s kappa = .76, percent agreement = 89%; see Table 1), and substantial reliability (Cohen’s kappa > .60) was reached for all categories (Pew Research, 2016; Viera & Garrett, 2005). Reliability on some features fell below a kappa of .60, likely because these features rarely appeared in the sample, leading to a highly skewed distribution. The researchers again discussed their discrepant coding until they reached agreement (Krippendorff, 2013), after which one researcher completed the rest of the sample (n = 74) (Macnamara, 2005).

We used Stata version 12 for all analyses. We produced descriptive statistics after which we performed bivariate analyses to examine the CAY Index scores by brand popularity (independent t-tests) and by alcohol type (one-way between-subjects analysis of variance [ANOVA]).

Ads from brands popular among youth did indeed have higher index scores (M = 12.8, SD = 3.6) than nonpopular brands (M = 10.7, SD = 3.9), t(94) = −3.28, p = .001. Popular and less popular brands had similar scores for production value and character appeals. Compared to less popular brands, popular brands used storytelling marginally more often, t(94) = −1.93, p = .056, and featured animals, t(69) = −2.04 p = .04, and minority actors more often, t(94) = −3.02, p < .01. Popular brands used marginally more youth-oriented genres, t(94) = −1.93, p = .06; used more of the reward features except sex and romance, though the difference was only significant for friendship, t(94) = −2.85, p < .01; had a marginally higher mean risk-related content score, t(94) = −1.96, p = .05; and used more depictions of inappropriate alcohol use, t(65) = −2.79, p < .01. Less popular brands used product appeals more frequently t(85) = 3.34, p < .01, and used quality, t(94) = 3.43, p < .001, and properties, t(94) = 2.15, p = .03, more often. See Table 3 for details.

A one-way between-subjects ANOVA revealed a significant difference in the use of appealing content by beverage type, F (2, 93) = 6.44, p = 0.002. Post hoc comparisons using the Tukey HSD test indicated that the mean CAY Index scores for both beer (M = 11.5, SD = 3.2) and liquor (M = 12.4, SD = 3.1) differed significantly from wine (M = 6.4, SD = 4.5). The CAY Index scores for beer and liquor did not differ significantly.

Compared to the other alcohol types, liquor ads featured more character appeals, t(63) = −2.17, p = .03; more animated characters, t(66) = −2.35, p = .02; more fantasy, t(51) = −2.07, p = .04; more sexual content, t(71) = −2.90, p < .01; and marginally more overconsumption, t(69) = −1.84, p = .07.

Beer ads used more males, t(63) = −3.87, p < .001, and more White actors, t(71) = −3.89, p < .001, than nonbeer beverages. Beer ads also featured more magic, t(73) = −2.22, p = .03, humor, t(84) = −3.22, p < .01, success, t(94) = −2.00, p = .04, and inappropriate use appeals, t(68) = −3.17, p < .01. Bud Light, the most consumed brand among youth (Siegel et al., 2013), used humor and storytelling in 83% of its ads.

Wine ads did not use any youth-oriented genres, but did use product appeals more frequently (M = 3.0, SD = 2.2) than beer (M = 1.9, SD = 1.6) or liquor (M = 1.1, SD = 1.0), F(2, 93) = 5.82, p < .01. Wine ads also had the highest average use of most of the product appeals, but only significantly so for benefits, t(45) = −2.69, p = .01, and properties, t(45) = −2.69, p = .01, compared to liquor ads. See Table 3 for differences by alcohol type.

Given the cross-sectional nature of the study, we are unable to state definitively that the greater use of youth-appealing content by popular brands is responsible for those brands being more popular among youth. We also cannot say whether the ads appeal to underage youth specifically or whether they appeal more broadly to nonyouth as well. Alcohol advertising is industry self-regulated in the United States, and according to the Distilled Spirits Council and the Beer Institute Codes of Advertising and Marketing (Beer Institute, 2011; Distilled Spirits Council of the United States [DISCUS], 2011), alcohol advertising cannot primarily appeal to underage youth. Future research should compare the associations between brand CAY Index scores and youth and adult brand consumption to determine if the appeals studied here are primarily appealing to youth over adults.

That said, as long as alcohol ads have content that appeals to youth at all, the presence of such ads during popular youth television shows indicates a need to reconsider current advertising placement policies (Jernigan, Ostroff, & Ross, 2005; Ross, Ostroff, & Jernigan, 2014). We further recommend that the alcohol industries revise their advertising codes to broaden the list of prohibited content beyond depictions of Santa Claus and underage models, and that they add specific examples of problematic content rather than just requesting that marketers avoid unspecified symbols, language, music, gestures, entertainers/celebrities, cartoon characters, groups (Beer Institute, 2011), depictions of “rites of passage,” and branding on items used by underage youth (DISCUS, 2011).

Despite its limitations, this study moves the literature forward by exploring how the content of alcohol advertising relates to adolescent interests and vulnerabilities, by providing an up-to-date and more comprehensive examination of the content creation strategies of alcohol companies, and by suggesting possible regulatory implications. The CAY Index is a unique tool for assessing the extent of youth-based appeals in alcohol advertising, and perhaps other substances such as tobacco or food and beverage advertising. Analyses using the index can help identify youth-focused marketing efforts and then inform media literacy programs and health communication campaigns aimed at countering these advertising messages.