Page 227 -

8.8 Conducting an interview 215

be perceived as confirming a hypothesis, is well known (Orne, 1962). If you hear
participants saying uniformly positive things about a system that you developed or a
model that you suggested, you might be a bit cautious about over-interpreting those
responses.
As with all research involving human subjects, interviews must be conducted in a
manner that respects the rights and concerns of the participants (see Chapter 15). Be
sure to clearly explain to your interviewees that they can decline to answer any ques-
tion. This is particularly important if you are discussing potentially sensitive topics.
When they do decline to answer, simply note their lack of response and move on to
the next question. Participants should have the chance to take breaks, particularly
if the interview is long. Interviews and focus groups should be kept to a reasonable
length—probably less than 2 hours (Brown, 1999).
Focus groups present additional challenges. Listening is still paramount, but you
may want to jump in to keep conversation on track. Focus groups can go badly wrong
in many different ways: discussions can digress; participants might talk at length to
the exclusion of others; disagreements might arise; or you might simply have a group
that doesn't get along well. If you see any signs of trouble, you can jump in, gently
urging participants to stay on topic, let others speak, be polite, and so on. You might
try to be particularly sensitive to participants who seem to be quietly observing with-
out saying much. Although some quiet folks might not have anything to say, others
might be intimidated. Particularly if your quieter participants appear to be agitated
or uncomfortable, you might address them directly, offering an opportunity to speak:
“Joan, is there anything you'd care to say about…?” Having asked this question, you
must be ready to accept “no, thanks” as an answer.

8.8.3.4 Promoting discussion
What if you hold an interview (or focus group) and nobody talks? Spending an hour
in a room with someone who responds in monosyllables is both unproductive and
unenjoyable. You certainly can't force anyone to talk in any detail, but you might
be able to encourage them. If your interview is fully structured you may not have
much wiggle room, but you can add questions to semistructured and unstructured
interviews, in the hopes of eliciting comments. If you are asking about user reaction
to a given tool, you might rephrase the question in various different ways—do they
use the tool at home or at work? Have they had problems at home or at work? Is the
spreadsheet tool good for personal finances and for taxes? In some cases, overly
general questions might discourage responses: if you dig deeper into specifics, you
might remind your interviewee of some specific incident or need that is relevant.
Physical props, note cards, and other probes can also stimulate feedback. In
Section 8.3.2, our sample exploratory questions included asking participants how
they currently arrange items in scrapbooks. Instead of simply asking this question,
you might give your interviewee a small pile of photos and ask him to arrange them
as if he was constructing a scrapbook, explaining the process as he went. This use
of probes can be particularly helpful for revealing attitudes and practices that your
participants may not have fully articulated: even though your interviewee knows