In the 1930s my dad, the psychiatrist J.L. Moreno, first developed social network analysis (a story I tell in my forthcoming book, Impromptu Man), he called it the sociology “of, by, and for the people.”
That wasn’t exactly the lesson learned when Facebook decided to conduct a study of the way its users’ moods were affected by more positive and more negative newsfeeds in 2012.
The social scientists associated with the Facebook study must not have counted on all the negative newsfeeds about their experiment that have popped up all over the web and focused the attention of just about every business news organization. Critics noted that the 2012 study of “emotional contagion” manipulated the positive and negative newsfeeds seen by 700,000 Facebook users and their assessed their responses without their specific consent (beyond the generic Facebook agreement). This was in effect a social psychological deception study that should have been reviewed like any other experiment; several academics were part of the research and it was published in an academic journal. The privacy and anonymity of the non-voluntary subjects is also an issue. The possibility that some Facebook-obsessives could have been harmed by the study would have been considered in any competent ethics review.
Facebook does engage in serious academic research about social networks using its data and it does in general let users know that if they are interested. In this case the idea of emotional contagion through social media posts was confirmed, which is useful knowledge. The point is not to suppress good social science about virtual interactions through increasingly ubiquitous social media networks but to do it in a way that avoids the impression that the companies that collect the data are an authoritarian corporate culture, the “malefactors of big data,” as Teddy Roosevelt might have said.
Why the outrage? Apart from the objections I’ve mentioned, I think the deeper answer lies in the popular notion that modern social media are neutral vessels, while in fact they are shaped and filtered by algorithms. Users can adjust their newsfeeds to avoid the kind of filtering that was done in the 2012 experiment, for example, by putting their setting on “most recent” rather than “top stories.”
That’s a maneuver that life doesn’t normally require of us, but if social network analysis is to be the people’s social science, as my dad intended, the people need to understand the way the networks are mediated.