Opinion: Learning as we go: Lessons from the publication of Facebook’s social-computing research

In the aftermath of the publication of the “emotional contagion” study conducted by Facebook researchers working with Cornell University scholars (1, 2), many observers weighed in about everything from the acceptability of undertaking the research at all, to how it was conducted, the rules and regulations that applied to it, and even the advisability of publishing the article. The views were disparate and conflicting (3⇓⇓–6). Our goal in this Opinion is not to try to settle these debates, but rather to attempt to draw some general lessons and offer recommendations from an ethics perspective as large-scale social-computing research moves forward. Our motivation is not to defend or chastise the research and technology communities or those responsible for ethics oversight of research. Instead, we wish to suggest that the development and application of an appropriate ethical framework and some form of ethics oversight is a moral imperative that is also in the interest of all. First, such oversight acts as a crucial signal of rules and accountability, which can increase overall trust and, in turn, the willingness to support and participate in research. Second, oversight leads to more credible research that others can build upon and funders and investors will support. Third, consistent approaches ease and encourage the trust and legitimacy needed for partnership and collaborations, which is the basis of 21st century science of all kinds.

The Facebook research example highlights areas of confusion and uncertainty with a growing category of investigation that is part innovation and part research, taking place in the context of very large user populations on Internet platforms that have become ingrained in the lives of users around the globe. We start with the presumption that large-scale social-computing research will continue, that it offers valuable insights, and that, like basic research …

↵¹To whom correspondence should be addressed. Email: jeffkahn{at}jhu.edu.