Social media platforms are now a prominent data source used by corporations, academics, and even the government to gain insights on our society.
A common tendency in the new era of big data is to quickly analyze every new dataset or tool and draw conclusions before fully understanding the implications behind it. In an articled titled, “Does Social Media Actually Reflect Reality?” Forbes contributor Kalev Leetaru argues that the downside of today’s social media analytics is that we rarely stop to question whether or not our results are truly meaningful.
There are many questions about the true representativeness of social media. For example, Tinder was the subject of numerous headlines after its CEO claimed that the service had “80 million users worldwide and 1.8 billion swipes per day.” The SEC filings, however, revealed that the company had only 9.6 million daily active users totaling 1.4 billion swipes. Even publicly traded companies provide conflicting information to the point where it becomes difficult to assess their accuracy.
Statistics drawn from social media can yield results vastly different from reality. Prior to the Iowa caucuses, Bernie Sanders led Hillary Clinton in Facebook mentions by a daunting 73%. Yet when the votes were tallied, they nearly tied with Hillary Clinton leading by a mere .2%. In the 2012 election, 9 million Facebook users utilized a new feature allowing them to tell their friends that they had voted. The data appeared to show that Obama had won a significant percentage of women voters, with nearly a two-to-one ratio compared to men. Upon exploring the data further, Facebook revealed that women are statistically more likely to share in general on Facebook. The skewed data was a result of the two-to-one ratio in Facebook sharing rather than a reflection of the election.