A new study from scientists at the McGill University and Carnegie Mellon University urges their academic peers to be wary of trying to predict human behavior using social media, according to an AJC.com report on Dec. 1.
The researchers note that thousands of papers are published every year that use social media data to inform and justify decisions and investments such as predicting how movies will do at the box office. However, McGill and Carnegie Mellon researchers warned different social media platforms attract different demographics, so it is difficult to get a handle on what the population as a whole is thinking.
One example they used to get their point across is that young female-dominated Pinterest wouldn’t necessarily be the best indicator on how well a primarily male-targeted movie will do. They also note a site’s design, like the absence of a ‘”dislike” button on Facebook, can cause some variance in results.
Another recent example of social media based research is a Facebook study done by the University of California and Cornell University that looked at how the social network impacted its user’s emotions. Facebook manipulated the newsfeeds of some of its users and researchers found those who were exposed to negative content had negative posts and vice versa. But, when the study came out, a lot of people were pretty mad Facebook had been messing with their newsfeeds.
AJC.com reported that researchers are not trying to discredit the social media studies, however, they are saying scientists need to find way to correct the biases that come along with the research in order to come up with more accurate results.
“The common thread in all these issues is the need for researchers to be more acutely aware of what they’re actually analyzing when working with social media data,” computer scientist Derek Ruths said in an interview.