Ethical research design has been a cornerstone of the human-computer interaction field since its inception. However, online research has increasingly challenged traditional ethical guidelines for human subjects research, especially with the widespread use of social media content and data. Academic and industry researchers alike, along with institutional review boards and other governing bodies, have struggled with how to deal with these new research paradigms in an ethical way. The goal of this workshop was to explore the difficult and unanswered ethical questions surrounding this tension, with an eye towards best practices for social media and big data research.
This workshop brought together researchers from both industry and academia in a multitude of subfields such as data security, behavioral health, gender, identity, law, and media studies. The overarching theme of presentations and discussions was that of the ethical challenges inherent in participants’ work around social media or big data, regardless of discipline. A number of short talks from participants framed the discussion. One theme was that of vulnerable populations: David Myles (University of Montreal) focused on privacy and data sensitivity while studying online mourning practices and Jessica Pater (Georgia Tech) talked about issues associated with studying presentations of behavioral health issues like eating disorders and self-harm across social media platforms. Several participants focused on the ethics of new movements within our research paradigm, including Anna Lauren Hoffmann’s (University of California, Berkeley) talk on justice and research ethics, Nicholas Proferes’ (University of Maryland College Park) discussion of frameworks for open data, Yew Chuan Ong’s (University of Sheffield) work on ethics and social bots, and a Twitter research case study from Sara O’Sullivan (University College Dublin). Finally, there were several talks regarding policies and standards for data ethics. Mutlu Binark (Hacattepe University) spoke about the movement in Turkey to establish data ethics policies, Michael Zimmer (University of Wisconsin Milwaukee) presented on conceptual gaps in research ethics policies, and Lauri Kanerva (Facebook) discussed how Facebook has been meeting these gaps and challenges with their newly established review policies. Casey Fiesler (University of Colorado Boulder) closed out the talks with a discussion of ACM SIGCHI proposals for reviewing procedures regarding ethics.
The group collected these insights, organized them, and synthesized commonalities across all genres of research and study that were represented in the workshop. Seven key open questions resulted from this discussion. First, do we need to redefine or make clear what we define as “public” as it relates to data? Second, what do we define as a human subject? We know that people are human subjects, information they share with us is human subjects data, but what about the information they share publicly online? Third, obtaining consent: if information shared online is human subjects data, are there situations where obtaining informed consent is appropriate or not? Fourth, should relational and associational harms associated with online data be a concern of our community? Fifth, is it acceptable for researchers to violate Terms of Service? If research necessitates breaking TOS, what are the additional research design oversights needed from an ethical standpoint to mitigate risk? Sixth, how do we account for users’ expectations in these discussions? Research has shown that tech savviness and digital literacy levels are inconsistent across populations, making this a very complex issue in and of itself. Finally, seventh, how do we address the use of stolen, leaked, or deleted data in our research? The participants were in agreement that this was the beginning of a much longer, nuanced, and needed conversation that we hope to see manifest within more thoughtful research designs, the addressing of ethical challenges and barriers within papers, and more open discussions within the social media research community and in future workshops.
Thank you to all participants for your contributions to this workshop and ongoing conversation! You can also find some tweets from the workshop at #icwsmethics. The workshop was moderated by organizers Casey Fiesler (University of Colorado Boulder), Anna Lauren Hoffmann (University of California, Berkeley), Jessica Pater (Georgia Institute of Technology), and Nicholas Proferes (University of Maryland College Park).
We anticipate organizing workshops in this area at a number of different conferences, and you can find more information about past and future workshops at sociotechethics.wordpress.com.