The AnthroGuys have been quiet but we’re still here, writing, moving, proposing, hiking, reading, researching, having our fall courses slashed, writing some more, and so on.

In the midst of all this I’ve had  something brewing about the relationship among data, insight/inspiration, and design — especially user-centered design.

First, I wonder if there is a false dichotomy brewing out there between design decisions based on “data” and those based on “insight.”  Second, I love how the data/insight issue mirrors some theoretical bugbears in anthropology and points to some affinities between good design and good anthropology.

Back in May, the New York Times ran a piece about Douglas Bowman, who left Google for Twitter, where he became creative director; his team is credited with adding the “trending topics” sidebar to your Twitter screen.  Apparently, Bowman left Google because, as the title of the Times article says, “data, not design, is king” there.  At Google, Web analytics rule the day and bold creative leaps are usually not welcome unless they are backed by solid data, which means that, in effect, they are not welcome.  Design Prof Debra Dunn of Stanford Institute of Design noted to the Times that Web analytics and related methods are good for some things (tweaking an existing design, or helping choose between option A and option B), but Web analytics do not produce broad design directions nor, typically, big leaps forward.  For that, Dunn said, one needs close-in engagement with users, understanding what they do and their pain points, and then some healthy design decisions can flow from that — decisions which can then be subjected to the Web analytics, but which cannot be inferred from Web analytics.  Bowman, who credits Google with doing what they do well (who doesn’t?), said nothing about user experience research per se, but noted that Twitter fits his sensibilities better because the organization is more open to inspirational leaps and design innovation.  (As far as I’m concerned, the trending topics bar is a super addition to Twitter, especially considering that they fielded it before #iranelection hit the mainstream news; at this writing, it’s still in the top 10).

The Times story stuck with me, perhaps because the “data is king” line about Google implies (perhaps unintentionally) that Web analytics generates data while Dunn’s suggested approach (going out and being with users, watching them, etc.) does not.  User experience blogger Andrew Hinton got me thinking even more with a thoughtful discussion of some of the same issues.

Hinton considers data to be both quantitative and qualitative, valuable and often essential, a great use in challenging ones own design biases, something clients often demand , BUT, not by itself the end of the story: “It’s just that data doesn’t do the job alone. We still need to do the work of interpretation, which requires challenging our presuppositions, blind spots and various biases.”  I love this quote from Hinton because it up-ends the positivist assumption that the data speaks for itself.  Data never speaks for itself, it always requires an act of interpretation (yes, even statistics are mute until we give them meaning!).  In design, the fact that data doesn’t speak for itself is especially obvious, since, as Hinton says, “Data cannot tell us, directly, how to design anything.”  What then should user experience professionals do?  According to Hinton, we should “use data to inform the fullest possible understanding of the behavior and context of potential users, as well as bring our own [design] experience and talent to the challenge.”  In other words, research-savvy designers need both data and designerly inspiration for good UX practice.

All of this reminds me of the anthropology as science vs. interpretive anthropology divide in my own discipline.  While explanatory science vs. interpretive understanding is not a necessary dichotomy, many have practiced anthropology as if it were.  On the science side, we have data (quant and qual), variables, causes-effect relationships, comparisons, and scalable conclusions.  The best work in this vein often leads to modest but reliable conclusions about human behavior.  On the interpretive side we have data (mostly qual), its interpretation, some insightful leaps, compelling illustrations, and a story that, if it’s well done, quite often takes us toward a deep understanding of another culture.

But wait!  A few decades of STS have shown us that science, like everything else man-made, only works via social and cultural means, and that capricious insights and idiosyncrasies — personal, social, cultural — matter greatly in how the work of science gets done.  Likewise, every piece of (good) interpretive anthropology begins with data, usually generated by fieldwork involving first hand, face to face contact with others.  So scientific and interpretive anthropology both involve data and inspiration.

Sometimes, the criticism is leveled at the interpretive folks that their findings are ultimately based on some mysterious leap of interpretation, and their conclusions are unverifiable and probably hinge on the ineffable qualities or skills of the researcher himself.  However, in my opinion, the masterpieces of ethnographic writing in anthropology have been produced by interpretive-leaning anthropologists, and they succeed in conveying some feel for what life is like in other cultures in a way that science-oriented anthropology often does not.  If deep understanding of another culture — flawed and open to debate, for sure — comes from leaps of inspiration and insight, then so be it.  The result can be beautiful (or well-designed, if you like).  The most recent work of well-designed interpretive anthropology that I read was Steven C. Caton’s Yemen Chronicle: An Anthropology of War and Mediation.  Caton has data, but he also takes liberties (in my opinion, warranted), and by the time you are done the people of whom he writes are less “subjects” and more “characters,” different but comprehensible.

No wonder design and anthropology go together so well.  In research-driven design work, data and insight are essential ingredients, just as they are in any good bit of anthropology.  If data and insight are in tension, it’s a productive tension — in both fields.

I close with a nice quote from Bonnie McDaniel Johnson in Design Research (edited by Brenda Laurel):

“Design research is inherently paradoxical: it is both imaginative and empirical.  It cannot be simply empirical because the ‘typical’ customers that researchers need to understand are rarely able to articulate their needs.  Design researchers must go beyond what they can find: to see more than is visible, and to learn more than can be heard.  Accordingly, design research is an act of imagination, just as much as design itself.  Yet it must also be grounded in empirical evidence, for no business manager wants to think that the research on which her profits depend is made up in the research department” (Laurel 2003:39).