Online social life is a poor predictor of reality

Posted Sep 14, 2017 by Tim Sandle
Social media platforms like Twitter are poor predictors of how people respond and feel in real-life, according to new research. Social media analytics have some way to go before they approach accuracy; something which affects marketing.
A woman shopping online
A woman shopping online
Keith Williamson, Flickr (CC BY 2.0)
Social media marketing is big business and this may work fine if a new television set is pitched to the right demographic. But for other services, such as pitching a certain type of entertainment experience or a welfare service, the data gathered may not be necessarily accurate. There are implications here for on-line advertisers, including both those who reach out to consumers and those that target other businesses. New research highlights the complex psychology of social media users.
The new research focuses on Twitter and the headline is that Twitter is an unreliable witness to the world's emotions; this arises, according to the University of Warwick, because online social life does not always reflect offline social reality.
According to sociologist Dr Eric Jensen there is no evidence that social media content shared on Twitter is a truthful reflection of how its users feel. This despite the flurry of people tweeting out 140 character messages on a platform where there are some 300 million active users each month. Dr. Jensen cautions that mining Twitter using big data analytics for meaningful social and psychological trends is likely to prove unfruitful.
This is because, the academic surmises, Twitter users have created, over time, their own unique cultural behavior. This includes the form that conversations take and social media personas that users adopt. In turn this cultural box shapes the way by which users present their views online. There are social conventions, deep power relationships and alternate identity influences, just as there are in real-world, but these take different forms online to the extent that analyzing a person’s Twitter activity does not necessarily lead to an accurate prediction of how that person will respond off-line.
Commenting on his findings, Dr Jensen notes: "Twitter users present only one side of themselves on social media, shielding their true feelings for good reasons, such as professional reputation.”
Moreover, drawing inferences from Twitter about what a particular social group think of an issue or predicting how they might respond to a product is flawed since Twitter users are not representative of the normal population. From Dr. Jensen’s analysis, Twitter users are more likely to be male and the rates that different people tweet at is skewed heavily towards more active users appearing in searches more often than less active ones, leading to the viewpoints of heavy tweeters dominating any information analysis.
The research has been published in the journal PLoS One, in a research paper titled “Putting the methodological brakes on claims to measure national happiness through Twitter: Methodological limitations in social media analytics.”