October 05, 2015
It turns out you can say a lot more in 140 characters than you think.
A study by the University of Pennsylvania that analyzed over 10 million posts from 5,000 Twitter users found that it was possible for computer scientists to guess a person's income based on their tweets, and a whole lot of other personal information as well.
“We can predict tons,” said Svitlana Volkova, a research assistant at Johns Hopkins who collaborated with UPenn, to the university's news service. Scientists from University London College also worked on the study.
The scientists didn't have access to the users' exact incomes, so they chose users who disclosed their job and used the average salary of that job as a proxy.
Researchers noticed some intriguing patterns. For example, high-income users tended to have more followers and talked more about politics, corporations, technology and justice.
In contrast, poorer users were more likely to use profanity and discuss personal issues.
"In general users of lower income use social media more for personal communication, while the ones with higher incomes use it for more ‘professional’ issues," noted the study.
Another finding: people who were openly Christian "earned significantly less on average than people who chose not to signal their religious belief," the study found.
The emotional content of a tweet also varied according to income. Richer users tended to exhibit more fear and anger while poorer users showed more anxiety.
"Users with higher income post less emotional (positive and negative) but more neutral content, exhibiting more anger and fear, but less surprise, sadness and disgust," the study found.
However, the researchers didn't find any correlation between higher incomes and higher levels of happiness or life satisfaction.
While there have been many studies on how social media can be used to discover a person's personality or political preferences, this is the first large-scale study on how to automatically guess someone's income level based on Twitter.
This information isn't just of use to scientists. Advertisers, insurance companies, politicians and potential employers are just some of the interest groups who will no doubt want to dig into the goldmine of personal information they can find in a person's Twitter feed.
Read the full study here.