Unstructured and Semi-Structured Interviewing
Interviewing is the most widely used method of data collection
There is a continuum of differing amounts of control exerted in
the different kinds of interviews.
Informal interviewing - lack of structure. "Just conversation" but
not easy! You must tax your memory and write up everything
ASAP before it fades. Good for establishing rapport.
Unstructured interviewing - based on a plan, but a with a minimum
of control. Good for getting people to "open up," and for when you
have plenty of time.
Semi-structured interviewing - here make use of an interview guide,
a written list of questions and topics to be covered; it is efficient and
allows for freedom and control at the same time.
Structured interviewing - when informants are asked to respond to
the same set of questions or stimuli (next unit). Most of you will be
doing structured face-to-face interviews with the people in your
sample. Asking the same questions allows you to talk about
frequencies and percentages--especially important if you have a
Starting up unstructured interviews:
Assure anonymity and confidentiality.
Explain that you want to know what THEY think.
Encourage them to interrupt.
Don't try to hide your intentions.
Let the informant lead after you have defined the focus;
and let him/her "take off" with supplying information.
Probing is a necessary and effective skill:
Silent probe - give informant time.
Echo probe - repeat last sentence, and lead further.
Uh-huh probe - use affirmative comments.
Long question probe - asking for more detail or
Be careful about leading questions - being directive or
putting words in the informants mouth: "I assume
your husband always discusses these thing with you,
Baiting probe- acting like you already know something
to get people to open up.
Set up practice interviews (not with friends) to iron out rough spots
and ambiguous questions.
A big problem with semi-structured and structured face-to-face
interviews: boredom and fatigue. Spread out the process over
time; take notes about the interview during the interview; describe
the setting (if different each time) and the reactions of the
Language skills are important.
Introduction: be cordial and non-judgmental; tell about the
purpose, anonymity, expected approx. length of the interview; be
Use a tape recorder if permission is given, but take notes at the
RESPONSE EFFECTS - measurable differences in
interview data that are predictable from characteristics
of respondents, interviewers, and environments, e.g.
middle-class interviewers get more conservative answers.
Deference effect - race, ethnicity, gender are significant
Threatening questions - may reduce accuracy or
willingness to respond.
Expectancy effect - creating through our subtle
behavior the objective results we want to see.
Be aware that people are often inaccurate reporters of
their own behavior.
You can try a cued recall (referring to records); or
aided recall (giving a list of possible answers); or
share a personal landmark. Keep response time
USE OF FOCUS GROUPS:
Group interviews of about 6-12 respondents with a moderator.
Can complement a survey or sample of structured interviews,
Not statistically valid, but can supply rich ethnographic data.
Requires a qualitative analysis and/or content analysis
(subjects to taken up later).