Of course, Twitter users are not a representative sample of all people--just who are the "twitter-afflicted" (as Adam Gopnik wrote in a recent issue of the New Yorker) who pick up their phone to tweet from a park? Still, Twitter users are a broad demographic, earlier research shows, and this approach to near-real-time remote sensing via Twitter posts--not based on self-reporting--gives a new window for scientists onto the shifting moods of very large groups.
Well, it is what it is. It's a complex subject and I acknowledge one could pick all sorts of holes in it if you wanted to.
> Of course, Twitter users are not a representative sample of all people--just who are the "twitter-afflicted" (as Adam Gopnik wrote in a recent issue of the New Yorker) who pick up their phone to tweet from a park? Still, Twitter users are a broad demographic, earlier research shows, and this approach to near-real-time remote sensing via Twitter posts--not based on self-reporting--gives a new window for scientists onto the shifting moods of very large groups.
Well, it is what it is. It's a complex subject and I acknowledge one could pick all sorts of holes in it if you wanted to.
"twitter study"