The College Alcohol Culture in Twitter: Developing a Methodological Approach

Theoretical Background and research questions/hypothesis:  High-risk drinking among students is associated with negative consequences impacting the individual student, other students, and the college community (Hingson, Heeren, Winter, & Wechsler, 2005; Perkins, 2002). The negative alcohol consequences for individual students may include blackouts, physical assaults, injuries, unintended and unprotected sexual activity, and alcohol poisoning (Hingson et al., 2009; Perkins, 2002). College football games are one example of event specific activities that are associated with high-risk drinking (Beets, 2009; Neighbors et al., 2011; Neighbors et al., 2007). Alcohol is referenced on the majority of public, social media profile pages of college students and underage drinkers (e.g., Egan & Moreno, 2011; Fournier & Clarke, 2011; Ridout, Campbell, & Ellis, 2012). Social media platforms like Twitter may provide accessible health related data by monitoring unique social interactions (Korda & Itani, 2013; Loss, Lindacher, & Curbach, 2014). Hashtags within Twitter create a conversation among users around a specific event. For example, #BostonStrong was created to react, cope, and express thoughts about the 2013 bombing event (Tournas, 2013). Another example is the use of hashtags by collegiate athletic programs to encourage dialogue and brand their programs (Cooper, 2010). To date, there is limited research related to the alcohol culture within Twitter, even though the advanced search tool provides an efficient mechanism to search tweets within hashtags. The purpose of this exploratory study is to determine the utility of collecting data from Twitter hashtags and propose a coding system for alcohol references in tweets.

Methods and Results (informing the conceptual analysis):  Among a group of 140 institutions of higher education (IHE), all but one had a Twitter account associated with the institution. All of the selected IHE had a Twitter account associated with the athletic department, and 51% had a football team Twitter account. Among the Twitter accounts, a total of 282 unique hashtags were found by searching Twitter, IHE websites, and by identifying the most used hashtag of the account. The advanced search function was used to search 10 alcohol types (i.e., beer, wine, bourbon, gin, rum, scotch, tequila, vodka, whiskey, liquor). Among the 9,197 tweets referencing an alcohol type, the majority, 58% included a reference to beer. To provide a nuanced approach to examining the college alcohol culture within hashtags, a codebook was developed to categorize the alcohol-related tweets. The tweets were categorized by activities associated with individuals: alcohol-related risk, alcohol-related consequences, and protective behavioral strategies. Protective behavioral strategies are actions that assist with the reduction of alcohol consumption and the negative consequences associated with alcohol use (Martens et al., 2005).

Conclusions:  Multiple search words can be used with the Twitter hashtag search to better understand the college alcohol culture. A codebook to categorize alcohol-related tweets could assist in assessing the college alcohol culture by determining if tweets are referencing alcohol-related risk, consequences of alcohol use, or protective strategies to reduce alcohol problems.

Implications for research and/or practice:  Examining hashtags may provide innovative approaches to acquiring accurate and time sensitive population level data to inform health educators about their institutions’ unique college alcohol culture.