SURVEY RESEARCHCHART 1 Interviewer Characteristics Can Affect Responses
EXAMPLE OF INTERVIEWER EXPECTATION EFFECTS
Female Respondent Reports That
Asked by Female Interviewer Whose Own Husband Buys Most Furniture
Husband buys most furniture 89%
Husband does not buy most furniture 15
EXAMPLE OF RACE OR ETHNIC APPEARANCE EFFECTS
PERCENTAGE ANSWERING YES TO
“Do you think there “Do you think that
are too many Jews in Jews have too
Interviewer government jobs?’’ much power?”
Looked Jewish with Jewish- 11.7% 5.8%
sounding name
Looked Jewish only 15.4 15.6
Non-Jewish appearance 21.2 24.3
Non-Jewish appearance and 19.5 21.4
non-Jewish-sounding name
Note:Racial stereotypes held by respondents can affect how they respond in interviews.
Source:Adapted from Interviewing in social researchby Herbert H. Hyman with William J. Cobb et al.;
foreword by Samuel A. Stouffer. © 1954, 1975 University of Chicago Press, p. 153.
The interview setting can affect answers. For
example, high school students answer differently
depending on whether we interview them at home
or at school. The presence of other people often
affects responses, so usually we do not want others
present.^67 For example, Zipp and Toth (2002) found
greater agreement on numerous attitude items when
a spouse was present at an interview; wives modi-
fied their answers to conform to their husbands’
responses and husbands’ changed little.
An interviewer’s visible physical characteris-
tics, including race and gender, can affect respon-
dent answers, especially for questions about issues
related to race or gender. For example, African
American and Hispanic American respondents
express different policy positions on race- or ethnic-
related issues depending on the apparent race or eth-
nicity of the interviewer. This occurs even with
telephone interviews when a respondent has clues
about the interviewer’s race or ethnicity. In general,
interviewers of the same racial-ethnic group get
more accurate answers than does an interviewer of
a different background. Gender also affects inter-
views both in terms of obvious issues, such as sex-
ual behavior, as well as support for gender-related
collective action or gender equality. Yet, as Weis-
berg (2006:61) noted, “Interviewer matching is
rarely used in the United States, except when it is
necessary to use interviewers who can speak
another language.... Interview matching is more
necessary in some other countries, as in Arab coun-
tries where it would be considered inappropriate for
an interviewer of one gender to speak with a respon-
dent of another gender.”^68
Interviewer characteristics can influence
answers in many ways. For example, when the inter-
viewer was a person with disabilities, respondents
lowered their self-reported level of “happiness”
compared to answering a self-administered ques-
tionnaire. Apparently, they did not want to sound
too well off compared to the interviewer. However,
when respondents completed a self-administered