12 Forbes Terrace
Pittsburgh, PA 15217












Wednesday, February 11, 2004

Rick E. Robinson
Rick E. Robinson, former chief experience officer at Sapient and founder of eLabs, visited Studio on Monday and gave a version of the talk he gave at Doors of Perception a few years back about ethnography and things to think with.

The object of research, he said, is to reveal the complex and pass on that information. There are three core parts to ethnography:

  • You go to them. It always happens in context.
  • You talk to them. Talk to the subjects, not read about them.
  • You write things down. Develop a disciplined set of data so that your findings can be passed along and used by others.

The research itself is comprised of four things:

  • A description. Of something: a thing, an activity, a belief, a setting, etc.
  • Interpretation. Not summary, not "insight," not wholly "emergent" either. It is grounded in the subject.
  • Towards an end. Research has to be both instrumental (useful to the people you do the research for) and salient (it has to be to the point).
  • Within constraints. Of site, setting, time, tools, material, solution spaces.

There isn't any one approach or method to ethnography: in fact, making up new methods is part of the interest in doing ethnographic research. You do, however, have to have a plan and go with a hypothesis that you can test. This way, you can engage with (and bring something to) the field; it keeps you motivated. A Hunt Statement is useful here. It details what you are going after, and teh best ones are this compound sentence: We are going after X so we can do Y.

There are simple heuristics that you can use to organize your observations, like the AEIOU (Actions, Environment, Interactions, Objects, Users) method and the Think/Do/Use method. There's also a broad range of techniques to gather data: guerilla research, cultural inventories, visual stories, beeper studies, visual diaries, video ethnography, interviews, etc. In a cultural inventory, you are trying to understand the vocabularies of perception: what is it that you see that allows you to create a context and understand cultural structures? Data comes from cultural production; in a cultural inventory, researchers are looking for underlying structures. In interviews, how people express their beliefs, attitude, and knowledge (ie how they tell their stories) is part of the data. Interview questions should be more open-ended than survey questions; you want them to construct stories.

For Rick, the real purpose of research is to create models of thought, which then become "things to think with." These models live between the setting of the research and what needs to be created. They describe something that is fundamentally "other" in a way that people who weren't there but have an interest in it can understand and apply it. Good models are like good art: subversive. Any representation or re-representation always offers the idea that things can be different. Examples of good models are Csikszentmihalyi's "Flow" to describe the optimal experience and Vygotsky's "Zone of Proximal Development."

When creating models, you are trying to make the research visible so that it can be commented on. The model itself doesn't have to be perfect, it just has to be right. Try to let go of the low-level data and just make a story, then slowly add data back in to make an argument for the story. The best models have a long explanatory reach, parsimony, communicative power, and a multi-disciplinary point of view.

posted at 08:27 AM in big ideas, special guest stars, techniques | comments (0) | trackback (0)


‹‹ preceding entries




3D (2)
alumni (3)
assistantships (3)
big ideas (23)
classes (16)
classmates (17)
cmu (12)
cognition (1)
cpid program (3)
design 101 (25)
design theory (20)
extracurricular (11)
faculty (8)
field trips (5)
hci program (4)
info design (3)
interface design (6)
meta (6)
money (5)
papers (4)
photography (5)
preparation (6)
projects (43)
readings (29)
software (8)
special guest stars (10)
student life (16)
teaching (1)
techniques (12)
thesis paper (4)
thesis project (1)
typography (7)
visualization (8)


Week of Mar 14, 2004
Week of Feb 29, 2004
Week of Feb 22, 2004
Week of Feb 15, 2004
Week of Feb 8, 2004
Week of Feb 1, 2004
Week of Jan 25, 2004
Week of Jan 18, 2004
Week of Jan 11, 2004
Week of Jan 4, 2004
Week of Dec 7, 2003
Week of Nov 30, 2003
Week of Nov 23, 2003
Week of Nov 16, 2003
Week of Nov 9, 2003
Week of Nov 2, 2003
Week of Oct 26, 2003
Week of Oct 19, 2003
Week of Oct 12, 2003
Week of Oct 5, 2003
Week of Sep 28, 2003
Week of Sep 21, 2003
Week of Sep 14, 2003
Week of Sep 7, 2003
Week of Aug 31, 2003
Week of Aug 24, 2003
Week of Aug 17, 2003
Week of Aug 3, 2003
Week of Jul 27, 2003
Week of Jul 20, 2003
Week of Jul 13, 2003
Week of Jul 6, 2003
Week of Jun 29, 2003
Week of Jun 22, 2003
Week of Jun 8, 2003
Week of Jun 1, 2003
Week of May 25, 2003
Week of May 18, 2003
Week of May 11, 2003


Full Entries


Want more spam? Sign up to receive this blog via email:


All straight lines circle sometime. - The Weakerthans