So Facebook conducted research via (according to the Wall st Journal) its internal ‘Data Science Team’ which is apparently not affiliated or endorsed by a third party research body or institution. Facebook used its existing terms of service as an ethical umbrella under which to operate, is seems, any research that any individual in the team wished to conduct:
‘”There’s no review process, per se,” said Andrew Ledvina, a Facebook data scientist from February 2012 to July 2013. “Anyone on that team could run a test,” Mr. Ledvina said. “They’re always trying to alter peoples’ behavior.”…Facebook said that since the study on emotions, it has implemented stricter guidelines on Data Science team research.’
(Reed Albergotti, Wall st Journal, July 2nd 2014).
From a researchers perspective it seems lackadaisical, irresponsible and ethically loose to have no ethical review and to rely solely on a ‘terms of service’ agreement that, although will have been ‘ticked’, may or (most probably) may not have been explicitly read. This, to me seems particularly irresponsible when considering the direct manipulation (not just the observance or analysis of) users’ feeds to enact an emotional response.
Ethical choices in online research
I have come across the assumption by researchers that the individuals interacting online understand explicitely the consequences and ramifications of publishing personal content. This deliberate oversight is invariably for the benefit of the researcher (commercial or otherwise).
Alyssa Richman (in Representing youth: methodological issues in critical youth studies by A.L. Best, 2007, NY University press) sees the public nature of online publication as an effective mitigation of ethical issues:
‘Lurking [observing online content without active participation: therefore ‘invisible’]can be an extremely valuable research technique; I would not have been able to collect the majority of my data without it…the participants…were unaware that their writings were being collected…this violation was mitigated by several factors: first, the public nature of the research spaces and, second, the publication aspects of bulletin-board postings.’
This researcher did not (unlike Facebook researchers) interact with the authours of the observed content, but I find it particularly interesting that by way of explanation for ethical decisions, the value and importance of the research is a foremost concern: the needs of the researcher are paramount.
An article by Nonnecke & Preece: Shedding Light on Lurkers in Online Communities(Journal article, 1999) accepts lurkers as a large (if amorphous and unidentified) part of the online ‘community’ whereas Richman (in Best, 2007, p197) states that ‘As outsiders, we [the researcher] must be even more attentive to ensure we are understanding the actual meanings’ (italics added).
Being outsiders – being separate – for Nonnecke and Preece is here an ontological impossibility: the separation of us and them in Richman’s observations is an artificial one and, I could suggest, is ultimately very helpful in justifying an ethical decision to separate, objectify and manipulate data without consent from its authours.
Another ontological point could be made from the assumed ability to ‘understand actual meanings’ which is certainly not a given in certain epistemologies.
Facebook data and scientism
Returing to our Facebook account (pun intended), reading articles of diverse opinion such as that of Duncan J. Watts: Stop complaining about the Facebook study (Guardian.com, 7th July 2014) and Why Facebook’s user manipulation research study is ethically troubling (by Ray Junco, professor at Iowa State Uni and Harvard Fellow, July 6th 2014, VentureBeat .com) are descriptive of a wider ethical argument. The search for quality data (either for the benefit of institutional wisdom or commercial profit) is enhanced by the volume of data that can be collected (increased without consent) – ethics are a hurdle to be removed or surmounted at the earliest opportunity. This argument is one that is enhanced by, and benefits from, the tenets of empiracle science, positivism and statistical analysis – approaches that are served by the volume and quality of raw, disembodied data:
‘more and better science is the best answer we have.’ (Duncan J. Watts, Guardian.com, 7th July 2014)
The argument for a more ethically cautious approach to online data collection and analysis is one that is supported by the acknowledgment of the importance and delicacy of personal and detailed data (anonymous or otherwise) that cannot so easily be separated from its origins or judged out of context. The individual is important, not necessarily in a humanistic sense but in the sense of individual experience and situation.
Duncan J. Watts (Guardian.com) suggests that this consentless analysis of increasingly large and publicly personal social data is part of inevitable technological progress and to let ethical considerations dog its progress is to prevent companies such as Facebook from fulfilling its mission statement of making our world a better place:
‘we should insist that companies like Facebook – and governments for that matter – perform and publish research on the effects of the decisions they’re already making on our behalf’ (ibid)
Duncan’s article is entitled ‘Stop complaining about the facebook study’ but I can only think of How I stopped worrying and learned to love the bomb… It sounds like Facebook is a de-facto online governing body, magistrates of our online affairs. Vote with your ‘likes’! It’s a juggernaught you can’t stop so why not get on board and help steer it to Babylon? Duncan J. Watts is, by the way, principal researcher for Microsoft Research dept…is he looking for a job at Facebook?
But many commentators are much more concerned with the why of the research – the purpose for which the data will be used: a going concern in a world of Wikileaks and Snowden revelations on big data and how it is used. To suggest that Facebook, Google, Microsoft are a benign force for good seems proposterous from almost every rational angle. And the idea that technology is always progressing for the benefit of mankind is a contested paradigm of the past and present that will stay with us into the future. Technology will save us!!!
Ethical issues concerning internet research are, as described here, by no means a simple process of ticking boxes on the ethics form of whichever insitution you happened to be associated with: commercial, educational or otherwise. Ethical concerns are neither simple, uncontested or fixed and the internet is a fast-developing interactive social space of active, shifting tectonic plate boundaries.
As a user, it seems useful to be wary of how your content is used. As a researcher, everything can be data but how and why is the data is being used? An ethical consent form does not provide you with the moral framework for your work: your ethical standpoint is something that is threaded throughout the fabric of research you are creating: as Facebook’s research team has learned – it will be visible from cyberspace.
More information on internet research ethics:
- Digital Ethnography
- British Psychological Society: Ethics of internet research
- ESRC ethics framework