Saturday, August 29, 2015

Using Twitter Data to Study Men and Women's Drinking Habits

My analysis of half a million drunk tweets, and more than two million tweets about alcohol, was just published by Quartz; you can read it here (and the coverage by Slate here, if you speak French).


  1. Hi Emma, big fan. I'm working on my epidemiology PhD at UNC, with a background in computer science. If you field comment-questions, can I ask:

    I've seen twitter data used more and more for some interesting effect - live tracking and response of food poisonings in Chicago, I believe, was one of the coolest. What do we know about twitter's demographics compared to, say, the US population, and (therefore) how generalizatable are these results? I know I've read twitter is a more black-friendly community than some social networks, for instance... curious who it is we're really talking about in these social media based and twitter specific studies.

    Best wishes!

    1. Hi Mike! Twitter data is, unsurprisingly, demographically skewed. Pew provides some good basic stats: Clearly, this doesn't make the data useless, but it does mean you should be mindful about how much your conclusions generalize.

  2. This comment has been removed by the author.