A Word of Caution about Facebook

Update, July 3, 2014: The Proceedings of the National Academy of Sciences has published an editorial expression of concern about the Facebook experiment conducted in January 2012.

Update, July 1, 2014: Forbes reports that the sentence mentioning research was actually added to Facebook's user agreement four months after the experiment occurred, and that the experiment may have included users under the age of 18.

StrangeScience.net began using Facebook to highlight new posts and related science news in 2009, for the most part without issue. But in June 2014, news broke that Facebook conducted an experiment in "emotional contagion" on nearly 700,000 of its users. The social network sent some users happy news feeds and other users sad news feeds, and noted users' subsequent posts, e.g., positive or negative emotional status messages. None of the people Facebook experimented on were explicitly informed of the experiment. The results of the experiment were then published in the Proceedings of the National Academy of Sciences (PNAS). The observed change was small — about 0.1 percent — and having one's news feed skewed to affect one's mood admittedly sounds like a First World problem. But this experiment is worrisome for several reasons.

Reason 1: This is not how you experiment on people. If you have ethics, you get informed consent. Facebook's actions may be technically legal as its user agreement says users consent to have their information used "for internal operations, including troubleshooting, data analysis, testing, research and service improvement." So the warning of experimentation is apparently a single word (research) buried in a user agreement longer than the (un-amended) U.S. Constitution. As news comedian John Oliver stated in his epic net neutrality rant, if you want to do something evil, hide it inside something boring.

Reason 2: Although the observed change was small across the entire study, the effects on individuals may have been significant. Maybe not. At this point, no one knows. One blog post wondered how many incidents of domestic abuse or road rage happened because of this experiment. My guess is: probably none. But the same post wondered about suicide attempts. That, too, may be a stretch, but I have friends who struggle with depression, and everybody has a tough time sometimes. Imagine having trouble at home, school or work, and seeing an everything-is-terrible news feed just because you're part of an experiment that nobody felt obligated to tell you about.

Reason 3: Having experimented before, Facebook may experiment again. Use Facebook at your own risk.

Reason 4: A reasonable person might wonder what all this experimentation is for. A big company wants to see how it can affect "emotional contagion" among thousands of people. What does that company want to try next?

Reason 5: Remember your crazy uncle, the one who never met a conspiracy theory he didn't like? This stunt kind of gives him a legitimate argument. Perhaps the worst thing about this whole episode is that PNAS essentially approved Facebook's shenanigans by publishing the study. In between maintaining StrangeScience.net (admittedly just a hobby site) and writing for NASA and NOAA, I've fielded my share of angry notes from conspiracy theorists — people who insist that evolution is a satanic plot and/or human-induced climate change is a communist plot, even people who believe in the chemtrail conspiracy (and holy crap, those people have issues). That a big corporation can experiment, no matter how insignificantly, on hundreds of thousands of people without their knowledge, and that the research can then be published in the journal of one of the most prestigious science academies on earth gives weight to the (usually wrong) notion that we're all being gas-lit all the time. And it makes it harder for ethical scientists and science communicators to do their jobs.

As of this writing (June 29, 2014), I have not decided to stop using Facebook, although that option isn't off the table. I'm waiting to see what happens next. If I come across news that Facebook has amended its user agreement and/or the National Academy of Sciences has issued a badly needed apology (an approximation of one finally issued July 3, 2014) for stamping its imprimatur on such a creepy experiment, I will update this page.

Michon Scott
Owner, StrangeScience.net

Back to Strange Science home

Home | Goof Gallery | Timeline | Biographies | Evolution | References | Search | Email

Bookmark and Share