The data collection is based on ethnography, but it falls between participant and non-participant observation. The researcher goes through a one-to-one contextual interview with the subject as (s)he works. The role of the researcher is similar to that of an apprentice to the subject's master. The researcher is more than observer in that (s)he must question what is happening and the "master" must talk through what is happening. Contextual Inquiry is based on the four principles of context, partnership, interpretation, and focus.
Context: To get as close as possible to physical presence. It allows the gathering of ongoing experience and concrete data.
Partnership: The role is actually more than apprentice-master. The researcher and the subject are collaborators in understanding the work. In particular, the interviewer/interviewee, expert/novice, and guest/host roles are to be avoided.
Interpretation: It is not sufficient to collect data. Interpretation is necessary to make a hypothesis about what the data means. If the method were being followed through to design, the hypothesis would have an implication for the design that could result in a design idea.
Focus: A clear focus steers the direction of the contextual inquiry and allows the researcher to keep the conversation on track. The project focus must be defined in advance. This allows the researcher to find suitable sites to visit, suitable people to talk to, and which sort of tasks to observe.
The structure of the contextual interview is as follows:
The conventional interview.
The first stage is to get to know the subject. This is helped by running the first stage of the contextual interview along conventional lines. The researcher introduces him/herself, explains focus, promises confidentiality, and asks for permission to record. This stage takes approximately 15 minutes.
This takes approximately 30 seconds and involves the interviewer explaining the rules for the contextual interview—that is, the customer works while the researcher watches and interrupts if there is something unclear or particularly interesting.
The contextual interview.
The subject works, the researcher watches, and, as an apprentice, asks questions, analyzes artefacts, makes notes, and drawings.
Using the notes made, the researcher goes over what has been observed and gives the customer a final chance to correct and expand on issues that were raised.
I used contextual interview with the four UK members of WWITMan. Contextual Design asks for a focus statement. This is a statement of what is being sought and implies what the interviewer should look for. The focus statement for the case study was: "How people share knowledge, learn, and solve problems in a community that operates in a distributed environment." Before the Contextual Interview proper, I had an unstructured interview with the manager of the group. The aim of this interview was to set the CoP in context by finding out:
the structure of the group,
its background and history,
its role in the wider context of the organisation,
the KM technologies in place in the organisation,
locations of community members, and their roles,
when people meet,
other groups with which people are involved, and
what problems are experienced.
The contextual interviews were with the four UK members of WWITMan. A whole day was spent with Wayne, one half-day session each with Dave and Mike, and two, separate half-day sessions with Stan. The contextual interview format was followed closely:
The short traditional interview to find out background information about the respondent.
The transition to explain the format of the contextual section.
The contextual section. The format for this was followed as per the prescribed method. The work was observed, questions were asked, and notes were made. However, the focus in Contextual Design is on the whole work picture, whereas the focus for this case study was rather more specific. This resulted in the collection of some data that turned out to be irrelevant for this particular study. This was a minor point, for the contextual interview yielded a large amount of relevant, rich data.
The wrap-up. In this case the wrap-up was also used to broach some areas that it was felt had not been satisfactorily covered. This generally turned out to be the respondent's use and impressions of communications media that had not been observed.