“Don’t drink and tweet”: ethical issues in social science research on social media

March 24, 2016

The Academy hosted a conference on the role of ethics in social media research on Monday, March 21 in London.

Falling on the 10th anniversary of the launch of Twitter, the conference featured a series of presentations from PhD students, early career researchers, academics and practitioners grappling with a range of ethical issues and questions social scientists have encountered in conducting research across social media channels. These included production, availability and control of online data, concerns over privacy and security, transformations in understandings of public and private spaces, and appropriate ways of engaging with youth.

The conference was sponsored in conjunction with New Social Media, New Social Science?, the Social Research Association, SAGE Publishing, and NatCen.

Professor Susan Halford, Director of Web Science Institute, University of Southampton, addressed some of the “disruptions” caused by social media to overarching ethical regimes. She said the scale and dynamism of data generated by the web contributed to a perfect storm, with a process of bureaucratising and organising ethics on the one hand being challenged by “data deluge” on the other.

Social media inevitably brings into question well established assumptions of traditional ethics regimes, such as who is generating and controlling data. “I don’t think most people think that when they’re tweeting they are producing data,” she said. “However much the intention is to be seen, it’s not about producing data for academic researchers.” It also raises questions of streamlining ethical standards across disciplines, as mathematicians, computer scientists and others who can use social media data will subscribe to different ethical regimes, causing confusion and undermining the credibility of ethics processes.

A recurring theme of the conference was the idea of “blurred lines” created by social media, transforming the methods, scope and length of research practices, such as notions of consent and accountability. If tweets are seen as “independent” rather than subjects of academic inquiry, ethics guidelines need to take into account what a participant expects to happen with his or her data. This raises questions of public versus private spheres, often seen as a binary, but in the online world more of a grey area, according to Ansgar Koene of the University of Nottingham.

“[Online] is a public sphere in the sense that people can look at it, but the way in which people are actually using it is more like a discourse between friends. You might think of it in terms of having a discussion in a public space, say a café or a pub. I accept that the person on the next table could probably listen in, but I don’t really expect them to use a recorder and record what I’m doing and then do an analysis of the discussion I had with my friends in the pub. So it’s not so clear cut to say that it’s publically accessible therefore I don’t need consent or I only need to ask the platform provider.”

Authorship and ultimate end-use of information have become more prominent in research, as “people do not have an intuitive sense about the privacy that should be expected from internet communication and how far things can go in this communication.” As a result, Koene said it’s not just a question of asking “do you consent or not, but what kinds of information do you think you need to have in order to do a real evaluation about giving consent?” One presenter pointed out that this requires consent to be an ongoing rather than a static process.

Elvira Perez Vallejos, Senior Research Fellow at the University of Nottingham, explored some of the ethical considerations when using data generated from children and the need for increasing digital literacy. This was being addressed by the iRights programme, empowering children to decide what is and isn’t safe, and enabling them to access the internet in more creative, knowledgeable, and secure ways.

She said far from being careless, children are acutely aware of risks associated with online activity. Koene also found this to be true, “which is one of the reasons [children] want to switch to things like snapchat because they are aware that they lose control over what’s been published or put online,” he said. “There is an invisible audience and you don’t know who it is that is seeing things on an open platform.”

Other presenters addressed the evolving nature of relationships between researchers and participants. One noted how communicating with participants through mobile applications like WhatsApp broke down the traditional divide, while other forms of engagement reinforced power imbalances. Another explained how she had to learn to be professional on twitter and not to see it as an extension of private life, advising others not to “drink and tweet.”

Issues around big data and data sharing prompted ethical discussions of public good and the extent to which social media is representative of society. One presenter pointed to the “fundamental issue of the information age: how do we design systems to make use of our data collectively to benefit society as a whole, while at the same time protecting people individually?” Another said that, although social media “looks democratic on the surface”, the idea that all voices are equal across these channels is an illusion.

The conference sought to debunk myths around a universal formula of ethics. Libby Bishop of the UK Data Service at the University of Essex said we must ask questions to assess which course of action develops moral virtues. “[Ethical guidelines] will not necessarily be consistent or free from contradictions. Almost certainly you’ll get different kinds of answers depending on what action you’re assessing with a specific framework,” she said.

An overarching principle, however, prevailed. As one presenter asked, “just because something can be done, does it mean it should be done?”