Wednesday 2 July 2014

Update: Facebook Experiments Lacked Support in User Agreement and May Have Included Minors


I posted yesterday about the ethical and possibly legal ramifications of the already infamous emotional manipulation experiment where researchers tweaked Facebook user feeds and studied ensuing user behaviour. The post yesterday gave links to a number of useful accounts and analyses, but I did also mention my own doubt that the research, including the subsequent publication of the findings, was covered by the Facebook user agreement. Today, Kashmir Hill at Forbes reveals that this is exactly what was the case. Not only did the user agreement not include "research", however, apparently Facebook realised what this meant for the defensibility of the study and retrospectively added "research" to the agreed to activities by users after the study's data collection in January 2012. In addition, it is also revealed that the study inclusion criteria did not exclude minors, and since Facebook allow users down to the age of 13, this means that the researchers may very well have been children without their or their parents' consent.

Both of these revelations are, of course, of substantial importance for the research ethical assessment of the study. Not least is the combination rather damaging not only for Facebook and its study leader Adam Kramer, but also for the non-Facebook employed researchers Jamie Guillory and Jeffrey Hancock. This since it may be assumed that the research ethical assessment that was allegedly performed at their universities, Cornell and the University of California, rested, at least partly, on the presumption of consent being implied by the Facebook user agreement. Moreover, this point is especially sensitive because of the possible enrollment of children, as research ethics standards, regardless of area, is especially adamant on rigorous consent procedures and protection mechanisms for children, as it is for other vulnerable groups, and mandatory involvement of their parents or guardians in one way or another, especially when they are below 15 years of age. Possibly, dirt may therefore spill over also on the journal PNAS's responsible editor Susan T. Fiske of Princeton University, whose responsibility it was to ensure the ethical soundness of the article before publication.

That's ethics. But, of course, today's revelation also means that there may be basis for substantial legal complaints. Not least, since Facebook and the involved universities are based in the USA – the heaven of civil lawsuits for astronomical amounts of money – it seems far from improbable that users who where included in the study may join in a class-action suit against (primarily) Facebook and the involved universities. Whether or not there would be grounds for administrative of criminal legal action is more difficult to assess, as I lack knowledge of sufficient details of the relevant sections of US law.


Tuesday 1 July 2014

Facebook Emotional Manipulation Experiment: A Collection of Readings

I will not make a real post of my own re the already infamous experiments (initially claimed to be military funded, but that, it seems, was a hoax), where Facebook allowed behavioural researchers to manipulate the allocation of status updates in personal feeds, to study the resulting emotional communicative behaviour of users. My own brief take is that, whatever else may be said on the matter, this is definitely not covered by the user agreement I've signed when joining Facebook. For while I did agree to Facebook testing out all sort of things to improve their service, I did certainly not agree to be a subject in a scientific research experiment, the result of which is published in a scientific journal. I also think that the study may harbour some substantive both methodological and research ethical difficulties, spilling over to not only Facebook, but also the prestigeous PNAS journal's editors, who seem to have taken proof of research ethical review rather lightly... But don't take my word for it, here are four selected sources, not all echoing my views exactly, which may help you make up your mind.

The first one simply set up what the whole thing is about in broad terms, providing a few useful links. The second discusses the scientific quality of the study, which is also important from a research ethics standpoint. The third is an account by a usually brilliant bioethics and research ethics law scholar, discussing the legal ramifications of the study, as well as details regarding what has and should have happened in procedural terms. The fourth is a purely research ethical account by a trusted bioethics colleague of mine. Enjoy!

1. Meyer, R: Everything We Know About Facebook's Secret Mood Manipulation Experiment, from The Atlantic.

2. Grohol, JM: Emotional Contagion on Facebook? More Like Bad Research Methods, from PsychCentral.

3. Meyer, MN: Everything You Need to Know About Facebook’s Controversial Emotion Experiment, from Wired.

4. Hunter, D: Consent and Ethics in Facebook’s Emotional Manipulation Study, from The Conversation.