Implications of ethnography for design
Apparently, Paul Dourish‘s paper on the implications of ethnography for design has caused quite a stir yesterday at the CHI conference. Since it is so topical to the themes of this blog, here is the abstract and the download link:
Abstract
Although ethnography has become a common approach in HCI research and design, considerable confusion still attends both ethnographic practice and the metrics by which it should be evaluated in HCI. Often, ethnography is seen as an approach to field investigation that can generate requirements for systems development; by that token, the major evaluative criterion for an ethnographic studies is the implications it can provide for design. Exploring the nature of ethnographic inquiry, this paper suggests that “implications for design†may not be the best metric for evaluation and may, indeed, fail to capture the value of ethnographic investigations.
Download paper (pdf, 324 kb, 10 pages)
What I most appreciate about Dourish’s paper is the perspective that ethnography is not a collection of research techniques, but rather a specific theoretical orientation to knowledge production and representation. In other words, ethnography is embedded in theory and thus the use of its common methodological techniques have significant theoretical implications for the outputs and assumptions of the study. The theoretical amnesia is the problem I have with “discount†ethnographic techniques, where it is applied without knowing the assumptions built into that technique.
The power struggle is that ethnography, and research in general, is seen has having a supporting role to design. Ethnography supports design decision-making, but does not make decisions itself or at least it comes into conflict when it tries to make decisions. This is the colonial role reversal that Dourish talks about where field anthropologists began to write the reports for colonial administrators. Ethnography came to support colonial decision making, which does not mean it endorsed colonial policies. As a supporting role, ethnography is thus “optional†as opposed to integral to design decision making. Yet, my favorite parts of the research process are planning and synthesis where you originally frame the questions (this is the theoretical part) and tie the data back into that theoretical framework. This is where ethnography is integral not optional, because if you frame the wrong problem you lose lots of time and money to correct things. And this is different from entering at the “lab portion†of the project when you are defining the product not what the problem is.
One of my favorite examples of ethnography as theory versus research technique is a card sort that I did for a large retail client to “redesign†their product categories. Card sorts are, of course, a common HCI technique to get to the “cognitive categories†by which people structure words/objects, etc. The outputs of my card sort were, of course, the architecture for the site in terms of product categories but also I had this meta-narrative about how people sorted products by gender categories and spatial configurations of the home and outside. While I took it for granted, the client partner pointed out to me how different my “results†were in a company that does lots of card sorts.
That meta-narrative was the result of my ethnographic philosophical orientation in which I, probably unconsciously, seek to represent the experience from the perspective of the people studied. Now from an output perspective, the product category hierarchy I developed was ethnographic as well (basically, it was just a product kinship diagram) and could eventually be measured on the site in terms of traveled paths, and shorter purchasing steps, etc. The meta-narrative was important and insightful to the client (having greater implications beyond the website), but was much harder to measure in terms of its implications for design. In fact, it told them more about how they should not design something as opposed to what they should do.
I can see the controversy this is causing because its a conflict over who makes decisions in design. In the interdisciplinary product development process, its tough to be the new kid on the block, even if you are not so new.