Wow, science can guess how much money you make based on your Twitter activity
It appears that Twitter isn’t simply for social purposes anymore: it can also be used to guess your income level, and we’re a little torn about whether this is mostly cool or mostly creepy.
Computer scientists at the University of Pennsylvania have conducted a study regarding how your Twitter activity might reveal a lot more than your age, gender, and political leanings. (Okay, so your Twitter does already reveal a lot about you, but this is still important.)
According to the Washington Post, Daniel Preotiuc-Pietro of Penn’s Positive Psychology Center led a team of researchers in analyzing over 10 million posts from over 5,000 public Twitter user profiles. The team gathered the data in August 2014 and utilized the most recent 3,000 posts for each user—which is making us think twice about our privacy settings.
The researchers, who also included Svitlana Volkova of Johns Hopkins University, Yoram Bachrach of Microsoft Research, and Vasileios Lampos and Nikolaos Aletras of University College London, then calculated the users’ income.
This part seems less than foolproof: it’s based on users’ self-described occupations on their profiles, and we know for a fact that sometimes those “occupations” are more aspirational than literal. Regardless, the researchers employed England’s job code system to sort these occupations and assign an average income for each code.
After that work was done, researchers used the data to determine whether there were links between an individual’s income, and the way they use Twitter.
Essentially, it seems that higher income and lower income users have different reasons behind why they use the site: high-income users, who also tend to have more followers, use Twitter for spreading information and tend to retweet more content. Lower income users, on the other hand, use Twitter more often for social purposes, and include more links in their posts. The reason behind this is unclear, but the correlation is definitely interesting.
Additionally, the researched analyzed the content of the tweets, and discovered that higher income users’ tweets inferred emotions of fear and anger more frequently than those of their lower-income counterparts. Maybe there’s some weight to the old adage that money can’t buy you happiness?
In all, it’s some food for thought, especially the next time you’re about to press ‘Tweet.’