My name is danah boyd and I'm a Principal Researcher at Microsoft Research and the founder/president of Data & Society. Buzzwords in my world include: privacy, context, youth culture, social media, big data. I use this blog to express random thoughts about whatever I'm thinking.

Relevant links:

Archive

Where HCI comes from (and where it might go)

I’ve been doing a lot of thinking about HCI (human-computer interaction) and my relationship to that field. I’ve been kinda frustrated with HCI. The name HCI implies that the field is about people’s relationship with machines and the interaction paradigms and designs that enable more efficient or enjoyable connections between the two. Many argue that this is the crux of my research. I’ve been resistant to this because I believe that I study human-human interaction that happens to have a mediated component to it.

This week, a new book appeared in my mailbox: HCI Remixed: Reflections on Works That Have Influenced the HCI Community (eds. Thomas Erickson and David McDonald). This book helped remind me that human-human interaction was part of HCI, even if the field seems not to emphasize that these days.

This book gave me all sorts of smiles. First, I’m a sucker for books of essays where I know half of the authors and drool with respect over the other half. Second, I love books that trace histories that I read long ago while offering fresh perspective and new contextualization. Third, I like books that challenge me to rethink my position on something. Through the perspective of contemporary HCI scholars, this book examines some of the core literature that is at the foundation of HCI and reflects on its relevance today. In walking down this memory lane, I was reminded of the many facets of HCI. There’s the HCI that’s about interfaces. There’s the HCI that’s about development processes, foundational to contemporary industry practice. There’s the HCI that’s about taking computation into the wild while also making it ubiquitous or invisible. There’s the HCI that’s about supporting collaboration and groups. All of these HCIs are in the history of HCI and it’s fun to read these eminent and emergent scholars reflect on the work done in all of these areas. This book made me long for the days when I felt like HCI was my home because it highlights a history that is still relevant to me. (Of course, some of what they discuss – Everett Rogers and Jane Jacobs, for example – goes beyond HCI.)

While this book has unbelievable breadth, my frustration with contemporary HCI often stems from my feeling that it has narrowed its focus over the years. While experimental psychology has been fully embraced by the field, many HCI scholars reject qualitative social science as irrelevant to HCI. There are plenty who embrace it, but the experimental psych approach dominates the conversations and work that does not follow the normative formula tends to not get published. Personally, I’m wary of most publications that make broad claims based on user studies with n=6 CS grad students. (While there are sound reasons for this methodology in certain subfields of psych, most of how it gets executed in HCI scholarship makes my toes curl.) As HCI tries to become a field in its own right, I feel increasingly alienated by it. I stopped going to CHI a few years ago because it no longer felt like my home (and the cost was way prohibitive). I stopped reviewing this year because I felt as though my criticisms were with the methodological approach of the field and thus I was doing a disservice to CHI.

Yet, HCI and its sister CSCW really were the beginnings of thinking about how people communicate in computer-mediated environments and it’s nice to see that history recounted. It’s nice to be reminded that qualitative work really was valued. Much of that seems to have been forgotten in an era of scholarship that requires user tests and design implications to be considered valid. What happened to work that focused on the interaction between humans and computers in the wild? Personally, I love work that analyzes how mega collective action by inhabitants of a system result in behaviors never predicted by the designers. This, unfortunately, doesn’t fit neatly into the build/test/explain cycle that dominates the field; thus, it tends to get published elsewhere. I hear things are changing and HCI is evolving in new ways, especially now that iSchools are starting to engage with the topic. Perhaps this book will remind more folks where HCI came from and open new doors for where it might go.

Print Friendly

7 comments to Where HCI comes from (and where it might go)

  • I might be a little out of touch on this, but I’d say that if you asked the best designers and design firms who are actually building great interfaces right now, they’ll tell you a few things:
    — they do not identify with HCI as a field or label; it has become associated with old-school academic approaches that are too far behind the times
    — many of them are doing qualitative research and using that research to design better interfaces and conceptualize new models…it’s probably more guerilla-style and less rigorous than “real” academic research, but gets the job done
    — “usability” might have replaced HCI in some circles as a label, but both have been superseded by user-centered design and “user experience” design; the difference being that you’ll figure out beforehand how to build what people want and need, rather than just fixing broken stuff after the fact
    — the lab for this stuff is the web…if the analysis of the research is bad, the products and interfaces you design will be bad and you’ll cease to exist on the web. Sink or swim.
    — HCI might still live on as a mandatory course in computer science programs, to try to get techies to think about how normal people might actually want to interact with systems, but I don’t think the techies tend to be the ones doing the user research, designs or usability testing in the real world anyway.

    Sorry, this is all just off the top of my head…I’d like to hear other reactions. Designers who have been paying attention to successful design firms like Adaptive Path (http://adaptivepath.com/ideas/essays/archives/) and User Interface Engineering (http://www.uie.com/articles/) just don’t seem to be very interested in the four-year-old academic papers being published now on HCI.

  • Your post is interesting and the book sounds great.

    The term HCI is used in more diverse ways that most terms I’ve encountered in academia/scholarship/research. So I think the HCI that you criticize is sort of a staw horse, in that it’s only one way the term is used and I don’t think the most popular way. I see at least four, relatively distinct, uses of the term in academia and industry.

    1) Computer Science. When the term is used, in my experience, it most commonly refers to HCI as applied in this area. Not only is it about computers interacting with Humans, but often doesn’t really have anything to do with humans. I was on an NSF review panel judging proposals to an NSF program called Human-Computer Interaction. The bulk of the proposals came from computer scientists and engineers and involved creating very cool interfaces, with no evaluation/examination of this interface with actual humans. No humans in most of the research at all. In this case, HCI, just means cool intefaces. (This is most consistent with my experience at ACM/CHI meeting and related journals). This also does not have anything to do with quantitative research as I know it.

    2) Experimental Psychology. This is what you speak about above, and this is my background and most of my research. Using traditional quantitative experimental methods, reducing phenomena into quantifiable components and systematically manipulating specific variables while holding others constant. Also, sometimes involves more elaborate structure equations etc., to explain non experimental, but quantifiable phenomenon. I see research similar to this in the relatively new, but exponentially growing, human-computer interaction group within the Association for Information Systems, especially at the Americas Conference for Info Systems.

    3) The usability gurus. This would be Jakob Nielsen and others, and their primary methodology is “guerrilla usability testing”, which utilizes small samples of people carrying out tasks utilizing various applications and systems. This method fits with usability professionals, and might be best represented by the Usability Professionals Association Conference. (By the way, when you mention a small sample of undergrad students I think you’re mixing together the experimental psychology and usability perspective. No self respecting experimental psychologist would ever generalize from such a small sample without adequate statistical power, since you need to apply inferential stats, though we certainly will generalize from undergrads to the world. In fact, Nielsen is clear in saying that usability testing is NOT “scientific research”. One of his alertboxes was all about the pitfalls of quantitative research.

    4) CSCW (for want of a better term). It’s interesting that I hear CSCW mentioned frequently on the internet researcher’s list. I did not realize there was that connection, tell I joined the list a year or so ago. As you probably know CSCW is largely qualitative and it’s difficult to get a CSCW paper accepted if it is quantiative. Best I can tell this community relies heavily on elaborate qualtiative methods, such as ethnography etc, which, of course, has its own type of rigour. Interestingly enough I learned about this community from a colleague who came from a computer science department and is now in the Ischool at Penn State.

    Through my career I have moved through many of these communities – all studying the same phenomena – with different methodologies, cultures, and philosophies

  • interesting. I just came across the term ‘human-computer-human interaction’ this week, and I liked it. but you yourself noted that there are many facets of HCI. perhaps many facets of ..HCHCHCH..? because there *is* HC, and CH, and HCH, etc. all valid. i.e. the point is not to shift wholesale from HCI to HCHI, but to more specifically articulate which facet of the interaction is being examined.

  • Xianhang Zhang

    Maybe we should coin the term [HC]{2,}I :).

  • I know someone who likes to talk about HCI by drawing a two people and a computer in a triangle with the two people at the bottom and the computer at the top. he then proceeds to talk about human-human, human-computer, computer-human, and human-computer-human interaction (though I don’t think he’s ever gone so far as to say h-c-h-c-h-…)

    in response to the original post: on the one hand, I feel your pain with respect to the demand for quantitative, clearly defined, easily comprehensible research results within HCI (and especially at CHI). on the other, I see an increasing amount of research that pushes against that traditional view, either indirectly or directly. I’m thinking here of, for example, work by Phoebe Sengers, Bill Gaver, and some of their colleagues/collaborators along the critical technical practice lines. while this is still somewhat fringe in the academic HCI community, it seems to be gaining ground, especially with the support and interest from industry practitioners in user experience, as noted above by Jeremy. I’m quite excited/interested to see how we as a community negotiate the use of methodologies that question or provide alternatives to an underlying philosophical commitment to objective evaluation. furthermore, I want to be an active participant in the discussions of how these methodologies, and their underlying epistemologies, get incorporated into our definitions of doing research and what counts as valid methods of knowledge construction.

    and yes, thank you for reminding me that I wanted to pick up a copy of HCI Remixed.

  • Brian O' Hanlon

    Danah, the computer history museum has a wonderful lecture about Economics of Distributed Computing, Jim Gray does some very logical maths work, to explain current models of computing and its various components. You’ll find it if you search Google video. Jim Gray’s ACM paper on the Sneaker Net was written a couple of years ago – well worth a read also, if you haven’t seen it – this is what Jim is thinking about more recently. I think it is interesting and makes sense. Computer History Museum also has a panel interview with the creator of the Palm Pilot – Jeff Hawkins etc. The interview is packed with great thoughts about human-computing interation. If you have ever heard Clay Christensen speak about business models, you will catch the various references used by the Palm Pilot founders too. It is funny, how Palm started out to be a small OS software provider, and by accident ended up in Hardware. When the funding was going in the order of popularity: Internet, Networking, Software, Hardware.

    B.

  • Brian O' Hanlon

    Is Hiroshi Ishii a contributor in the book?

    Wrote a quick little summary of the lecture here:
    http://designcomment.blogspot.com/

    B.