UPDATE 30 June 2014, 8:00pm ET: Since posting this, Cornell has updated their press release to state that the Army did not fund the Facebook study. Moreover, Cornell has released a statement clarifying that their IRB
concluded that [the authors from Cornell were] not directly engaged in human research and that no review by the Cornell Human Research Protection Program was required.
Where this leaves the study, I’m not sure. But clearly something is amiss: we’re still sans ethical oversight, but now with added misinformation.
So there’s a lot of news flying around at the moment about the study “Experimental evidence of massive-scale emotional contagion through social networks,” also known as That Facebook Study. Questions are being asked about the ethics of the study; while I want to post a bit more on that issue later, a couple of facts for those following along.
Chris Levesque pointed me to a Cornell University press release noting that the study in question received funding from the US Army Research Office. That means the study did receive federal funding; receipt of federal funding comes with a requirement of ethics oversight, and compliance with the Common Rule. It is also worth noting that the US Army Research Office has their own guidelines for research involving human subjects:
Research using human subjects may not begin until the U.S. Army Surgeon General’s Human Subjects Research Review Board (HSRRB) approves the protocol [Article 13, Agency Specific Requirements]
Unless otherwise provided for in this grant, the recipient is expressly forbidden to use or subcontract or subgrant for the use of human subjects in any manner whatsoever [Article 30, “General Terms and Conditions for Grant Awards to For-Profit Organizations“]
I’ve also been in touch with Susan Fiske, the editor of the study. Apparently, the Institutional Review Board (IRB) that approved the work is Cornell’s IRB. That IRB found the study to be ethical:
on the grounds that Facebook filters user news feeds all the time, per the user agreement. Thus, it fits everyday experiences for users, even if they do not often consider the nature of Facebook’s systematic interventions. The Cornell IRB considered it a pre-existing dataset because [Facebook] continually creates these interventions, as allowed by the user agreement (Personal Communication, Fiske, 2014).*
So, there’s some clarification.
Still, I can’t buy the Cornell IRB’s justification, at least on Fiske’s recounting. Manipulating a user’s timeline with the express purpose of changing the user’s mental state is, to me, a far cry from business as usual. Moreover, I’m really hesitant to call an updating Facebook feed a “pre-existing dataset.” Finally, better people than I have talked about the lack of justification the Facebook user agreement provides.
This information, I hope, clarifies a couple of outstanding issues in the debate so far. Personally, I’d still like to see a lot more information about the kind of oversight this study received, and more details on the Cornell IRB’s analysis.
* Professor Fiske gave her consent to be quoted in this post.
Michelle Meyer argues that no institutional review was required, because the two coauthors helped design the research and interpret data, but did not perform the experiment itself. http://www.thefacultylounge.org/2014/06/how-an-irb-could-have-legitimately-approved-the-facebook-experimentand-why-that-may-be-a-good-thing.html
Pingback: Treating like cases alike: bioethics edition | The Broken Spoke