Jan Chipchase has some great thoughts on how to do more effective field work. He does a good job of explaining some of the experiences I’ve had working on my own projects. We ran into similar situations working on the NSYR. In the first wave of data collection we interviewed ~260 teens but found serious diminishing returns. On the second wave, we only interviewed ~130 and were already seeing a lot of overlap. Of course, these were all U.S. teens being interviewd by other Americans. Does it pay to do more interviewing and spend more field time when the researcher is not from the country under study?

I’m also a huge believer in the constant data processing approach that Jan advocates, though I’ve never had the luxury of dedicated data manager. This works for all kinds of research, not just ethnographic work in the field. Even when Peter and I were doing quick and dirty lab tests at Epinions, we would schedule time before and after every session to process what we had just learned and what we were learning overall. This happened more informally during the NSYR interviewing but it was still critical. The research team (yes, there should always be a team) needs to be processing field notes, photos, etc. constantly to ensure that the analysis can take advantage of the subtle, contextual information that isn’t captured on the video/audio recorder. If you wait days or weeks to do some initial processing, many of the important information that drove the use of contextual research in the first place will be lost.

Of course, it is extremely important to be wary of biasing your analysis too much in a particular direction. One particularly striking interview/session can lead you to grab onto a theme or thought that will heavily bias your interpretation of later ones. But that’s the fun and challenge of field work and research in general. It’s a subtle back and forth between what the world is telling you and what you are telling the world.

Update: Anne Galloway also did some thinking based on Jan’s post.