This week seems to be the week Facebook owns up to its faults, or at least to some of it. On Wednesday, Facebook apologized to the drag community for its discriminatory "real names" policy. On Thursday, the social network says it will revise its research policy over the backlash that ensued over the 2012 emotion experiment that surfaced in June.
Facebook chief technology officer Mike Schroepfer admits that "it is clear now that there are things we should have done differently," referring to the experiment done on nearly 700,000 users without their knowledge and consent where researchers tweaked their news feeds to elicit positive or negative emotions from them. Schroepfer, however, stopped shy of issuing a direct apology.
"We should have considered other non-experimental ways to do this research," he says. "The research would also have benefited from more extensive review by a wider and more senior group of people. Last, in releasing the study, we failed to communicate clearly why and how we did it."
Schroepfer says it has created new research guidelines with the help of academics. Part of these guidelines now require all studies to undergo an "enhanced review process" before research and "further review" if the research involves members of the academic community. The review will be conducted by a panel comprising Facebook's "most senior-area researchers" and members of the social network's research, engineering, legal, policy and privacy teams. Schroepfer also says all new engineers going through Facebook's six-week training program will also be trained on research guidelines.
While critics who have panned Facebook for its handling of the emotion experiment say this is a laudable first step to improving Facebook's research practices, others questioned its efforts. For example, Zeynep Tufekci, a fellow at Princeton University's Center for Information Technology Policy, criticizes the lack of a more objective review panel since Facebook did not choose to include outsiders in the review process.
"What if there is a conflict?" asks [subscription required] Tufekci. "Are there different standards for research that won't be published?"
Cornell University professor Jeffrey T. Hancock, who was also one of the authors of the controversial emotion study, says he is "very encouraged" by Facebook's decision to create a review board but that Facebook still needs to disclose its research guidelines in full to obtain feedback from the user and academic community and ensure that any research it conducts does not violate users' rights.
"Will they keep doing those and not publish them?" Hancock says, referring to experiments similar to the emotion study. "Or does the review panel say we need to think about that? They don't say anything about informed consent or debriefing."
Facebook's career page shows an opening for a Research Management Lead who will be responsible for "maintaining our research training and compliance programs" and will "ensure our policies are informed by best practices from the appropriate research communities."