We have recently presented our research at the 2014 Australasian Digital Humanities conference. This was an ideal audience to air our work in front of as they understand both the humanities aspect of the research along with the developing methodological concepts and concerns behind its design. As you might imagine, we were very interested to engage in discussion with this group about the mobile health aspect of our research so far.
The discussion was, as expected, great. The second question picked up on the crumb that we dropped in the conclusion which is, is this research ethical? The premise is just because we can access this information, does that necessarily make it OK? Moreover, individual users may understand that when they post to the Twitter platform, they are doing so publicly. However, they are most likely unaware researchers are mapping their activity and understanding how their participation relates to a broader conversation and network.
The first question from the audience was what was surprising in the research and how has it contributed to the scholarship.
The surprising element of the research has been the clear division between individuals and consumer groups or lobbyists within the conversation. Although they are probably equally distributed, they are also clearly separate from each other within the network. One group of users are genuinely interested in the conversation for what we can hypothesise to be personal interest, while the other group are interested in the policy discussion because of the impact policy has on their business/special interest group – a ‘typical’ policy interest group.
As for the research’s contribution to the scholarship, I thought it was appropriate to pick up on the provocations of the conference, specifically utopian research is necessary to understand what is possible. Yes big data research has been critically examined and many flaws highlighted within its many methodological applications. We even highlighted some of them in the presentation. However, this type of research highlights the new forms of knowledge that are possible within social network analysis. But yes, big data methods do deserve to be critically examined in light of how conclusions have been drawn.
The method has also made us realise how important it is to have ethnographers and computer scientists working on the same problem from the get go. The expertise and languages of each discipline can be vastly different, so they need to be calibrated early on in the project to extract the most important findings.
The second question was, how did we achieve ethics approval for this research? The short answer was we didn’t. When I mentioned this, the room was clearly divided on our approach, and I noticed physical scoffs and eye rolls for our work. Does this make it less important? Are we rogue researchers for conducting this project? It would seem that half of the room thought so.
My response was, we could easily have said on an application form that the information is public, and that’s that. The research is for the most part ethically approved. However, what we are suggesting is that the work does not necessarily fit within this framework – what we are also asking is that more regulation is not the answer. So what is the delicate balance here?
Indeed one person suggested, in a rather blunt comment, that this is not too dissimilar to marketers close reading the White Page to cold call potential clients. This is true, however I would argue that our rationale is slightly more engaging than economic return. We are conducting this research to better align industry practice with user cultures – a socially constructed approach.
Tim Highfield and Tama Leaver are also interested in this approach and have been conducting similar research on the Instagram platform. They presented research which also asked similar questions and we can draw on some of their conclusions to help answer these ethical questions.
Firstly, the social media contradiction (Leaver and Lloyd, 2014) which distinguishes between a user’s understanding of social interaction on social media platforms, and the provider’s analytical interrogation of the media itself. Social media contradiction then, highlights the problem succinctly to explain the conundrum we face as researchers.
Secondly, researchers need to weigh intentionality of sharing into the equation (Highfield and Leaver, 2014). The argument that just because it’s publicly available does not necessarily mean it’s fair game. A more unique approach to understanding participation is to understand the user’s intention.
This more than points to the process of ethnography as the logical next method to shore up our claims that have been highlighted through big data social network analysis. Ethnography will highlight not only intentionality, but also the difference between why the user participates and how we read that participation as researchers.