"Hey Everyone, I'm Drunk." An Evaluation of Drinking-Related Twitter Chatter

J Stud Alcohol Drugs. 2015 Jul;76(4):635-43. doi: 10.15288/jsad.2015.76.635.

Abstract

Objective: The promotion of drinking behaviors correlates with increased drinking behaviors and intent to drink, especially when peers are the promotion source. Similarly, online displays of peer drinking behaviors have been described as a potential type of peer pressure that might lead to alcohol misuse when the peers to whom individuals feel attached value such behaviors. Social media messages about drinking behaviors on Twitter (a popular social media platform among young people) are common but understudied. In response, and given that drinking alcohol is a widespread activity among young people, we examined Twitter chatter about drinking.

Method: Tweets containing alcohol- or drinking-related keywords were collected from March 13 to April 11, 2014. We assessed a random sample (n = 5,000) of the most influential Tweets for sentiment, theme, and source.

Results: Most alcohol-related Tweets reflected a positive sentiment toward alcohol use, with pro-alcohol Tweets outnumbering anti-alcohol Tweets by a factor of more than 10. The most common themes of pro-drinking Tweets included references to frequent or heavy drinking behaviors and wanting/needing/planning to drink alcohol. The most common sources of pro-alcohol Tweets were organic (i.e., noncommercial).

Conclusions: Our findings highlight the need for online prevention messages about drinking to counter the strong pro-alcohol presence on Twitter. However, to enhance the impact of anti-drinking messages on Twitter, it may be prudent for such Tweets to be sent by individuals who are widely followed on Twitter and during times when heavy drinking is more likely to occur (i.e., weekends, holidays).

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Alcohol Drinking / epidemiology*
  • Alcoholic Intoxication / epidemiology*
  • Humans
  • Social Media / statistics & numerical data*