Yesterday, the Journal of Medical Internet Research Public Health and Surveillance published findings of a year-long study conducted by researchers at the University of Utah. The National Institutes of Health gave researchers $721,825 in federal funding to harness geotagged tweets about food from over 603,363 unique Twitter users to glean larger regional trends about happiness and health in the United States. (Ah, yes, the occasional boons of inhabiting a surveillance state.)
For the purposes of the study, researchers compiled a list of 1,430 popular food words from the US Department of Agriculture’s National Nutrient Database. They labeled fruits, vegetables, nuts, and lean proteins as "healthy foods" while slotting such phrases as "McDonald's" and "Kentucky Fried Chicken" under fast food. The algorithm was prone to imperfections—for example, tweets about Steph Curry, husband of Ayesha, were initially counted under food given his surname—but over the course of the year, researchers collected and processed 4,041,521 geotagged food tweets that they mapped against 2010 census data.
The study found that the most tweeted-about foods included coffee, beer, and pizza, followed by "Starbucks," IPA, wine, chicken, BBQ, ice cream, and tacos. (Items one may expect to be talked about often on social— bacon, chocolate, donuts—didn't appear until much later down the list.) The most tweeted-about fast food joint, Starbucks, was trailed by Chipotle, Taco Bell and Buffalo Wild Wings. Tweets about chicken made up most of the tweets about healthy foods, followed by eggs, salad, turkey, bananas, and mangoes. Urban areas corresponded to more tweets about fast food, while tweets from economically disadvantaged neighborhoods didn't mention healthy foods as often.
There's a nifty visualization of all of this in the form of a map on GitHub. It's a compelling study, of course, though gleaning larger trends from aggregate Twitter data has its obvious limits that have a lot to do with our existing habits on social—namely, what we choose to consciously broadcast to the world. "The content of tweets reflects the type of information that people feel comfortable reporting," the study warns, "and may not represent the true spectrum of their feelings or their experiences. For instance, people may feel most comfortable presenting a neutral stance rather than voicing polarizing viewpoints."
Are you on Twitter? What foods do you find yourself tweeting about? Let us know in the comments!