User-generated alcohol-related content on social media. Determinants and relation to offline alcohol use

Alcohol is the most commonly used psychoactive substance worldwide. It is considered to increase social cohesion, and to play a part both in the development and expression of identity. Alcohol is, however, also associated with a range of adverse effects for the individual, and involves a great financial and social burden for society. Social media has currently become very popular commodity for social bonding, identity formation and expression, and social influence. This thesis consists of three papers which are based on survey-data, and examines determinants and consequences of user-generated alcohol-related content on social media within a student population. Paper 1 and 2 have a cross-sectional design, while paper 3 has a longitudinal design. Paper 1 investigated students’ disclosure of alcohol on social media, and identified characteristics related to frequent disclosure. The result from paper 1 suggested that most students had posted alcohol content on social media, but few students reported to do so frequently. The alcohol-related content shared primarily portrayed positive aspects of alcohol use. Several characteristics, beside alcohol use, were associated with frequent disclosures, which suggest that disclosure of alcohol content on social media is not solely determined by alcohol use. Younger, single, extroverted students, with lower openness scores were for instance more likely to report frequent disclosures, compared to their counterparts. These characteristics may be related to an increased interest in presenting oneself as social and an increased tendency to adhere to social norms. Frequent disclosure was also associated with more lenient alcohol-related cognitions. Frequent disclosure had the strongest association to reports of frequent exposure to similar content, which suggest that perceived norms for social media behaviour may be the most important determinant of alcohol disclosure on social media. Paper 2 focused on exposure to alcohol on social media. The main aim of this paper was to investigate characteristics associated with reporting frequent exposure. The results from paper 2 suggest that the level of exposure to alcohol content on social media is considerable in the student population. Alcohol-related content on social media was most frequently interpreted as depictions of positive aspects of alcohol. Several factors predicted exposure. Number of online-friends and frequency of logins to social media were positively associated with exposure to alcohol-related content, which seems self-evident as more time spent on social media and more possible senders of social media content are likely to result in more exposure to all types of social media content. Traits that have been associated with an increased tendency to observe or be interested in others’ (i.e. lower extroversion scores, and higher agreeableness and self-monitoring scores) were associated with frequent exposure to alcohol-related content, which suggest that the amount of exposure to alcohol on social media is partly determined by the level of attentiveness towards others in general. Frequent exposure was associated with high alcohol use and lenient alcohol-related cognitions, which could suggest that level of exposure is determined by attentional biases toward alcohol and friendship selection processes. The crosssectional design employed prevents however conclusions regarding directionality. In paper 3 longitudinal data was used to examine the relationship between frequent disclosure/exposure at time 1 and alcohol use at time 2 (one year after). The main research question in paper 3 was whether disclosure/exposure may indicate heightened alcohol use over time. A second aim was to identify factors that may explain the relationship between disclosure/exposure and alcohol use. The findings from paper 3 clearly suggested that frequent disclosure and frequent exposure to alcohol-related social media content can indicate heightened alcohol use. Controlling for potential confounding variables (e.g. demographics and personality factors, social media use, and alcohol-related cognitions), weakened the association between disclosure/exposure and later alcohol use, which suggest that such factors may partly explain the relationship between frequent disclosure/exposure and high alcohol use. Controlling for baseline alcohol use (time 1) resulted in the strongest weakening of the relationship between alcohol-related disclosures/exposures and later alcohol use, which could suggest that the relationship between disclosure/exposure and later alcohol use could largely be explained by the association between concurrent alcohol use and alcohol disclosure/exposure on social media. Disclosure of content depicting positive aspects of alcohol use predicted later alcohol use, even when all covariates were controlled for. This finding could suggest that positive alcohol disclosures on social media may cause a sustained or increased alcohol use, or that the link between positive alcohol disclosures and later alcohol use is explained by other common factors not included in the study. In summary, the thesis’s results suggest that students occasionally disclose alcohol-related content on social media, but that they are more frequently exposed to such content. The participants reported to both share and see content referring to positive aspects of alcohol more often than content referring to negative aspects. The current results suggest that high alcohol use and lenient alcohol-related cognitions are important determinants of disclosure and exposure. In addition, disclosure may be driven by self-presentation and norm adherence intentions, while exposure may be determined by the level of interest in others and engagement in social media. Disclosure and exposure of alcohol-related content indicates heightened alcohol use over time, this effect seems largely to be explained by disclosure/exposure reflecting alcohol use. The predictive value of disclosure/exposure on later alcohol use was dependent on the type of content shared or seen. Disclosure of content referring to positive aspects of alcohol predicted a slightly increasing or sustained alcohol use even when all covariates were controlled for, which suggest that disclosure of this type of content may yield a small causal effect on later alcohol use