Alcohol Portrayals on Social Media (Social Media)
The depiction of alcohol is the focus of a growing number of content analyses in the field of social media research and entails the phenomenon of alcohol brands and social media influencers marketing alcohol as well as ordinary social media users providing alcohol-related self-presentations.
Abstract
The depiction of alcohol is the focus of a growing number of content analyses in the field of social media research. Typically, the occurrence and nature of alcohol representations are coded to measure the prevalence, normalization, or even glorification of alcohol and its consumption on different social media platforms (Moreno et al., 2016; Westgate & Holliday, 2016) and smartphone apps (Ghassemlou et al., 2020). But social media platforms and smartphone apps also play a role in the prevention of alcohol abuse when they disseminate messages about alcohol risks and foster harm reduction, abstinence, and sobriety (Davey, 2021; Döring & Holz, 2021; Tamersoy et al., 2015; Westgate & Holliday, 2016). Field of application/theoretical foundation: Social Cognitive Theory (SCT; Bandura 1986, 2009) as the dominant media effects theory in communication science, is applicable and widely applied to social media representations of alcohol: According to SCT, positive media portayals of alcohol and attractive role models consuming alcohol can influence the audience’s relation to alcohol. That’s why positive alcohol portayals in the media are considered a public health threat as they can foster increased and risky alcohol consumption among media users in general and young people in particular. The negative health impact predicted by SCT depends on different aspects of alcohol portrayals on social media that have been traditionally coded in manual content analyses (Beullens & Schepers, 2013; Mayrhofer & Naderer, 2019; Moreno et al., 2010) and most recently by studies relying on computational methods for content analysis (e.g. Ricard & Hassanpour, 2021). Core aspects of alcohol representations on social media are: a) the type of communicator / creator of alcohol-related social media content, b) the overall valence of the alcohol portrayal, c) the people consuming alcohol, d) the alcohol consumption behaviors, e) the social contexts of alcohol consumption, f) the types and brands of consumed alcohol, g) the consequences of alcohol consumption, and h) alcohol-related consumer protection messages in alcohol marketing (Moreno et al., 2016; Westgate & Holliday, 2016). For example, a normalizing portrayal shows alcohol consumption as a regular and normal behavior of diverse people in different contexts, while a glorifying portrayal shows alcohol consumption as a behavior that is strongly related to positive effects such as having fun, enjoying social community, feeling sexy, happy, and carefree (Griffiths & Casswell, 2011). While criticism of glorifying alcohol portrayals in entertainment media (e.g., music videos; Cranwell et al., 2015), television (e.g., Barker et al., 2021), and advertising (e.g., Curtis et al., 2018; Stautz et al., 2016) has a long tradition, the concern about alcohol representations on social media is relatively new and entails the phenomenon of alcohol brands and social media influencers marketing alcohol (Critchlow & Moodie, 2022; Turnwald et al., 2022) as well as ordinary social media users providing alcohol-related self-presentations (e.g., showing themselves partying and drinking; Boyle et al., 2016). Such alcohol-related self-presentations might elicit even stronger identification and imitation effects among social media audiences compared to regular advertising (Griffiths & Casswell, 2011). Because of its psychological and health impact, alcohol-related social media content – and alcohol marketing in particular – is also an issue of legal regulation. The World Health Organization states that “Europe is the heaviest-drinking region in the world” and strongly advocates for bans or at least stricter regulations of alcohol marketing both offline and online (WHO, 2020, p. 1). At the same time, the WHO points to the problem of clearly differentiating between alcohol marketing and other types of alcohol representations on social media. Apart from normalizing and glorifying alcohol portayals, there are also anti-alcohol posts and comments on social media. They usually point to the health risks of alcohol consumption and the dangers of alcohol addiction and, hence, try to foster harm reduction, abstincence and sobriety. While such negative alcohol portayals populate different social media platforms, an in-depth investigation of the spread, scope and content of anti-alcohol messages on social media is largely missing (Davey, 2021; Döring & Holz, 2021; Tamersoy et al., 2015). References/combination with other methods of data collection: Manual and computational content analyses of alcohol representations on social media platforms can be complemented by qualitative interview and quantitative survey data addressing alcohol-related beliefs and behaviors collected from social media users who a) create and publish alcohol-related social media content and/or b) are exposed to or actively search for and follow alcohol-related social media content (e.g., Ricard & Hassanpour, 2021; Strowger & Braitman, 2022). Furthermore, experimental studies are helpful to directly measure how different alcohol-related social media posts and comments are perceived and evaluated by recipients and if and how they can affect their alcohol-related thoughts, feelings, and behaviors (Noel, 2021). Such social media experiments can build on respective mass media experiments (e.g., Mayrhofer & Naderer, 2019). Insights from content analyses help to select or create appropriate stimuli for such experiments. Last but not least, to evaluate the effectiveness of alcohol marketing regulations, social media content analyses conducted within a longitudinal or trend study design (including measurements before and after new regulations came into effect) should be preferred over cross-sectional studies (e.g., Chapoton et al., 2020). Example Studies for Manual Content Analyses: Coding Material Measure Operationalization (excerpt) Reliability Source a) Creators of alcohol-related social media content Extensive explorations on Facebook, Instagram and TikTok Creators of alcohol-related social media content on Facebook, Instagram and TikTok Polytomous variable “Type of content creator” (1: alcohol industry; 2: media organization/media professional; 3: health organization/health professional; 4: social media influencer; 5: ordinary social media user; 6: other) Not available Döring & Tröger (2018) Döring & Holz (2021) b) Valence of alcohol-related social media content N = 3 015 Facebook comments N = 100 TikTok videos Valence of alcohol-related social media content (posts or comments) Binary variable “Valence of alcohol-related social media content” (1: positive/pro-alcohol sentiment; 2: negative/anti-alcohol sentiment) Cohen’s Kappa average of .72 for all alcohol-related variables in codebook* Döring & Holz (2021) *Russell et al. (2021) c) People consuming alcohol N = 160 Facebook profiles (profile pictures, personal photos, and text) Portrayal of people consuming alcohol on Facebook profiles Binary variable “Number of persons on picture” (1: alone; 2: with others) Cohen’s Kappa > .90 Beullens & Schepers (2013) d) Alcohol consumption behaviors N = 160 Facebook profiles (profile pictures, personal photos, and text) Type of depicted alcohol use/consumption Polytomous variable “Type of depicted alcohol use/consumption” (1: explicit use such as depiction of person drinking alcohol; 2: implicit use such as depiction of alcohol bottle on table; 3: alcohol logo only) Cohen’s Kappa = .89 Beullens & Schepers (2013) N = 100 TikTok videos Multiple alcoholic drinks consumed per person Binary variable “Multiple alcoholic drinks consumed per person” as opposed to having only one drink or no drink per person (1: present; 2: not present) Cohen’s Kappa average of .72 for all alcohol-related variables in codebook Russell et al. (2021) N = 100 TikTok videos Alcohol intoxication Binary variable “Alcohol intoxication” (1: present; 2: not present) Cohen’s Kappa average of .72 for all alcohol-related variables in codebook Russell et al. (2021) N = 4 800 alcohol-related Tweets Alcohol mentioned in combination with other substance use Binary variable “Alcohol mentioned in combination with tobacco, marijuana, or other drugs” (1: yes; 2: no) Cohen’s Kappa median of .73 for all pro-drinking variables in codebook Cavazos-Rehg et al. (2015) e) Social contexts of alcohol consumption N = 192 Facebook and Instagram profiles (profile pictures, personal photos, and text) Portrayal of social evaluative contexts of alcohol consumption on Facebook and Instagram profiles Polytomous variable “Social evaluative context” (1: negative context such as someone looking disapprovingly at a drunk person; 2: neutral context such as no explicit judgment or emotion is shown; 3: positive context such as people laughing and toasting with alcoholic drinks) Cohen’s Kappa ranging from .68 to .91 for all variables in codebook Hendriks et al. (2018), based on previous work by Beullens & Schepers (2013) N = 51 episodes with a total of N = 1 895 scenes of the American adolescent drama series “The OC” Portrayal of situational contexts of alcohol consumption in scenes of a TV series Polytomous variable “Setting of alcohol consumption” (1: at home; 2: at adult / youth party; 3: in a bar; 4: at work; 5: at other public place) Polytomous variable “Reason of alcohol consumption” (1: celebrating/partying; 2: habit; 3: stress relief; 4: social facilitation) Cohen’s Kappa for setting of alcohol consumption .90 Cohen’s Kappa for reason of alcohol consumption .71 Van den Bulck et al. (2008) f) Types and brands of consumed alcohol N = 17 800 posts of Instagram influencers and related comments Portrayal of different alcohol types and alcohol brands in Instagram posts Polytomous variable “Alcohol type” (1: wine; 2: beer; 3: cocktails; 4: spirits; 5: non-alcoholic d