National substance use patterns on Twitter
Authors:
Open Access
- 6 November 2017
- journal article
- research article
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 12 (11) , e0187691
Abstract
We examined openly shared substance-related tweets to estimate prevalent sentiment around substance use and identify popular substance use activities. Additionally, we investigated associations between substance-related tweets and business characteristics and demographics at the zip code level. A total of 79,848,992 tweets were collected from 48 states in the continental United States from April 2015-March 2016 through the Twitter API, of which 688,757 were identified as being related to substance use. We implemented a machine learning algorithm (maximum entropy text classifier) to estimate sentiment score for each tweet. Zip code level summaries of substance use tweets were created and merged with the 2013 Zip Code Business Patterns and 2010 US Census Data. Quality control analyses with a random subset of tweets yielded excellent agreement rates between computer generated and manually generated labels: 97%, 88%, 86%, 75% for underage engagement in substance use, alcohol, drug, and smoking tweets, respectively. Overall, 34.1% of all substance-related tweets were classified as happy. Alcohol was the most frequently tweeted substance, followed by marijuana. Regression results suggested more convenience stores in a zip code were associated with higher percentages of tweets about alcohol. Larger zip code population size and higher percentages of African Americans and Hispanics were associated with fewer tweets about substance use and underage engagement. Zip code economic disadvantage was associated with fewer alcohol tweets but more drug tweets. The patterns in substance use mentions on Twitter differ by zip code economic and demographic characteristics. Online discussions have great potential to glorify and normalize risky behaviors. Health promotion and underage substance prevention efforts may include interactive social media campaigns to counter the social modeling of risky behaviors.Keywords
Funding Information
- National Institute of Environmental Health Sciences (5K01ES025433)
This publication has 48 references indexed in Scilit:
- An Exploration of Social Circles and Prescription Drug Abuse Through TwitterJournal of Medical Internet Research, 2013
- Using Twitter to Examine Smoking Behavior and Perceptions of Emerging Tobacco ProductsJournal of Medical Internet Research, 2013
- A Comparison of Peer Influence Measures as Predictors of Smoking Among Predominately Hispanic/Latino High School AdolescentsJournal of Adolescent Health, 2013
- Prevalence and correlates for nonmedical use of prescription opioids among urban and rural residentsDrug and Alcohol Dependence, 2013
- The Neighborhood Alcohol Environment and At-Risk Drinking Among African-AmericansAlcohol, Clinical and Experimental Research, 2011
- Pandemics in the Age of Twitter: Content Analysis of Tweets during the 2009 H1N1 OutbreakPLOS ONE, 2010
- Neighborhood smoking norms modify the relation between collective efficacy and smoking behaviorDrug and Alcohol Dependence, 2009
- Who needs liquor stores when parents will do? The importance of social sources of alcohol among young urban teensPreventive Medicine, 2007
- Predictors of Alcohol and Drug Use: A Multi-Ethnic ComparisonThe Journal of Social Psychology, 1995
- The theory of planned behaviorOrganizational Behavior and Human Decision Processes, 1991