SURVEY RESEARCHProbe A follow-up question asked by an interviewer
to elicit an appropriate response when a respondent’s
answer is unclear or incomplete.prepared for reactions such as “How did you pick
me?” “What good will this do?” “I don’t know
about this.” “What’s this about, anyway?” The inter-
viewer explains why a specific respondent, not a
substitute, must be interviewed.
The interview’s main part consists of asking
questions and recording answers. In a standard
interview (not conversational), the interviewer uses
the exact wording on the questionnaire, adds or
omits no words, does not rephrase, and asks
questions in order without returning to or skipping
questions. He or she goes at a comfortable pace and
gives nondirective feedback to maintain interest.
In addition to asking questions, the interviewer
accurately records answers. This is easy for closed-
ended questions, for which interviewers just mark
the correct box. For open-ended questions, the inter-
viewer’s job is more difficult. He or she listens care-
fully, must write legibly, and must record what is
said verbatim without correcting grammar or slang.
More important, the interviewer never summarizes
or paraphrases. Doing so causes a loss of infor-
mation or distorts answers. For example, the
respondent says, “I’m really concerned about my
daughter’s heart problem. She’s only 10 years old
and already she has trouble climbing stairs. I don’t
know what she’ll do when she gets older. Heart sur-
gery is too risky for her and it costs so much. She’ll
have to learn to live with it.” If the interviewer
writes, “concerned about daughter’s health,” much
is lost.
The interviewer knows how and when to use a
probe, a neutral request to clarify an ambiguous
answer, to complete an incomplete answer, or to
obtain a relevant response. Interviewers recognize
an irrelevant or inaccurate answer and use probes
as needed.^64 There are many types of probes. A
3- to 5-second pause is often effective. Nonverbal
communication (e.g., tilt of head, raised eyebrows,
or eye contact) also works well. The interviewer can
repeat the question or repeat the reply and then
pause. She or he can ask a neutral question, such as
“Any other reasons?” “Can you tell me more about
that?” “How do you mean that?” “Could you
explain more for me?” (see Expansion Box 13,
Example of Probes and Recording Full Responses
to Closed Questions).
Respondents often interpret straightforward
questions differently than the survey designer
intended. For example, “Inaccurate reporting is
not a response tendency or a predisposition to be
untruthful. Individuals who are truthful on one occa-
sion or in response to particular questions may not
be truthful at other times or to other questions”
(Wentworth, 1993:130).
Techniques to reduce misunderstanding, such
as conversational interviewing, deviate from the
standardized interview model. Beyond concerns
about introducing bias, conversational interviewing
requires more time and more intense interviewer
training. Yet as Conrad and Schober (2000:20)
have observed, respondent “comprehension can
be made more consistent—and responses more
comparable—when certain interviewer behaviors
(discussions about the meaning of questions) are
lessconsistent.” Paradoxically, nonstandardized
interviewing can increase reliability by improving
the consistency in how respondents interpret the
meaning of survey questions and responses.
Given this complexity and possible distortion,
what should the diligent survey researcher do? We
should at least supplement closed-ended question-
naires with open-ended questions and probes.
Open-ended questions take more time, require
better-trained interviewers, and produce responses
that may be less standardized and more difficult to
quantify. Fixed-answer questionnaires based on the
naïve assumption model imply a more simple and
mechanical way of responding than occurs in many
situations. The inquiry into interviewer bias, cul-
tural meanings, and the interview as a social situa-
tion provides a lesson in how qualitative and
quantitative styles of social research complement
one another. In all research we strive to eliminate
sources of interviewer bias and respondent confu-
sion. In the past decade quantitative survey
researchers discovered that qualitative researchers
offer valuable insights into how people construct
meaning in diverse social settings.