Why is Data Science Overlooked in Psychology?
And more importantly, why is psychology overlooked in data science?
This year, I attended some events for the largest science festivals in the world; Pint of Science. On the whole, the events themselves were excellent; though this article isn’t designed to be a review, I particularly enjoyed an evening of ‘Can Computer Games Make the World a Better Place?’. I already knew about some of the tech on show that evening (specifically, the stereo-training games for those with Amblyopia), but there was still a great deal to learn. And I love learning. The speakers were highly entertaining and, as someone with a vested interest in the digital, extremely useful for broadening horizons. (In short, props to the organisers of the event for hosting such a fantastic and informative discussion)
One speaker, however, grabbed my attention the most; Professor Steve Benford. He spoke of his research with a Japanese performance art group in order to use data science in games involving people. As you can guess, I found this deeply intriguing. The team from the computer science department developed their research around the paths people take in a collaborative online game; with one subject guiding another to locations (both complete strangers to each other) with the use of optionally-phrased cues and clues. Results of this are useful in many ways; numbers of clues issued, the GPS pathing people decided to take, etc. As any good talk should, it left me with many solid questions.
At the end of the discussion about his career, I asked Prof. Benford if they had a) collected any behavioural data throughout the various pieces of research, b) performed any statistical analysis on said potential data, and c) collaborated with anyone from a field such as my own in order to bring more results to bear. It is safe to say that in all points I was disappointed; sadly, the research team had done none of these things. He did mention that “that side of things” was interesting, and could have potential for addition to research in the future, but I very much got the impression that this was not any key concern. And, why should it be? Behavioural research is not his field and anything outside of what pertained to the actual research hypotheses would be wasting time and money.
It did get me thinking, however, about interdisciplinary research. It is a field I am rather interested in myself, but often become frustrated by as it is, naturally, implicitly niche. Some of the lecturers I have encountered through my course have been less digitally competent than many of the students — why, then, should they be interested in collaborative research with either data or computer science?
During my undergraduate final year, I voluntarily opted to take a module outside of my department; something that seemed to confuse any academic I spoke to within Psychology. Indeed, the convenor of the module (who I had to meet with in order to gain permission to undertake the course) was also inherently confused. I think at one point he looked blankly at me and said, “But why? Why do you want to study ethics in computing?”. My answer at the time, of course, was simple; this was the only course within the computer science department that I was allowed to take. I am nowhere near technically-educated in their field enough to participate in anything else, sadly. However, there was another aspect to my response that I was reluctant to divulge — for it is has never been understood by those I have spoken to.
It seems that for an undergraduate, interdisciplinary research isn’t the done thing. Despite IBM’s recent campaign that pushes the idea that we are entering the cognitive era of computing, nobody (at least within the small academic circle I am exposed to) seems to want to pick up on this encroaching boom. As another interesting fact, it seems much easier for those to down-specialize; as in, it is easier for those of a higher technical level to specialize in a field that is less technical than their own. This is for obvious reasons. It is much more difficult for someone to acquire practical programming ability (and experience) from a social science background than for someone who is an engineer in robotics to acquire knowledge of the human mind. This means that those on ‘my side of the fence’ (i.e., of the first example) are perhaps much more disinclined to try.
But then, perhaps those that are interested in collaboration between fields are part of the spearhead. Coincidentally, the computing ethics module that I undertook resulted in the highest grade I have received in my entire degree. Naturally, this was because I was highly advantaged at writing essays on ethics, due to my many years doing so in psychology — but I still take this as a sign that getting into interdisciplinary research is a plausible career path. Data science and psychology, from my perspective, go hand in hand. Though it is not what I wish to focus on in my own research, I have in the past been offered opportunities at my university’s own Horizon Digital Economy Research Institute, which is just beginning to delve into the connections between data and how humans work (particularly socially). At least in this instance, interdisciplinary research between areas of computing and psychology seem to be becoming more popular or commonplace.
As any honest writer acknowledges, my own perspective is a very small view of the world. There are thousands of departments and researchers willing to take part, and even “specialise” (if such a term can be appropriately used) in collaborative research, interdepartmentally. I am currently reading the pop-science text Mind Change, by House of Lords peer, Baroness and somewhat-questionable author Susan Greenfield. Her work, and the work she discusses in the book, is prime for collaborative research. Though I disagree with her on many points within the text and without (See: Autism and Internet Use), the general area of the discussion is interesting. It is, at least in spirit, nice to see that others take an interest in crossing the realms of psychology with both digital environments and technologies.
Sadly, the other side of the fence still exists. Fervently.
As one Sarah Byrne wrote for The Guardian two years ago, and another Claire Shaw wrote three years ago, interdisciplinary research is often plagued by barriers. Without even considering the new potential problems that the United Kingdom’s attempt at Brexit may cause for scientific research, the governing powers that be are not always as enthusiastic about such collaboration as they claim to be. Both articles openly discuss a review by the Higher Education Funding Council for England, which (whilst seemingly positive) does contain the addendum that the Research Excellence Framework (REF) has a quantitatively lower inclusion of interdisciplinary work in its outputs between 2009 and 2013. The first article I cited also mentions the trend of “publishing or perishing”, which may fuel this reluctance to rate interdisciplinary work highly as papers may “not have the necessary impact in any single scientific field to be worth publishing”. I would be inclined to believe this trend may not be as robust under the influence of emerging technologies (ironically), as self-publishing, pre-publishing and open-source science become more popular and less taboo. Still, it is an issue that science as a whole will have to resolve if interdisciplinary work is to flourish more.
There are also problems for interdisciplinary research that many do not actually put into words. Felicity Callard and Des Fitzgerald made an well-constructed argument, also for The Guardian, late last year that put things relatively well in perspective. When collaboration comes down to including social and psychological sciences, this may be where some scientists falter. There is always the problem of the nature of these two fields (something that I will not discuss in depth here, as I’m writing a whooole separate piece on that) and how the more changable, qualitative nature of their data comes into conflict with ‘purer’ sciences. The piece by Callard and Fitzgerald is worth reading, however — as is (as far as I can tell from a light read) their free, open-source book on the same topic. I feel as though I will soon likely ditch Greenfield for this latter text out of preference against biased digital criticism (as those whom have read Mind Change may — or may not — agree).
Despite these troubles, then; is interdisciplinary research even worth bothering pushing for? Nature, one of the more influential journals on the planet, seems to think so. The National Science Foundation in Virginia also paints it in what is clearly a positive light. Many prominent universities, in the UK at least, are creating their own interdisciplinary research departments in order to specifically facilitate such research. It also seems common sense — at least to a follower of science and believer in its usefulness to humanity — that more research is better than less.
Whether the scientific and academic communities can currently handle the crossing of the streams, however, is an entirely disputable matter. I would like to believe so, as I have outlined my personal experience of wanting to develop this area myself and having performed well in it. I hope, for my own future if not for the good of mankind — that the obstacles and problems outlined will eventually be overcome. Perhaps academia has a lot to learn in itself, and a lot of room for internal development, before an environment that caters to all walks of study can be cultivated.
