Full text: Brückner, Hannah: Surveys don't lie, people do?

Butler (1976) report that interviewers who obtained high ratings for interviewing skills from their 
team leaders produced actually more interviewer variance than those rated less skilled. One study 
found that more experienced interviewers used more unprogrammed speech and frequently altered 
the wording of questions (Bradburn 1983). Bradburn (1983, p. 314) hence concludes that flexibility 
of interviewers in adjusting their behavior and speech to the specific interview situation is a "real asset 
in carrying out surveys of the general population." 
While mainstream methodological wisdom deals with sources of bias by eliminating the 
interviewer as an other," interaction theory predicts (1) that this does not yield unbiased data, and 
(2) it yields biases which we cannot easily identify as such. Argyris (1980) points out that the 
unilateral control exercised over the subjects of rigorous research may lead them to become either 
overly sensitive to researchers' behavior and try to please or second-guess them, or to develop 
uncooperative and "negativistic" attitudes towards the task and play games of minimal involvement 
(p.48-49). In contrast, Brenner (1978) argues that in the interviewing situation subjects are free to use 
their own relational standards and deviate from the task. Interviewers, however, are in a paradoxical 
structural position: they should "make the interview act without acting" (p. 138). Specifically, they 
have to remedy interactional trouble in the interview without getting involved with the substance of 
the reason for trouble since that might influence respondents answers. When reference to formal, task, 
and etiquette standards fails, interviewers have "to give in to the idiosyncratic constructions of the 
respondent, that is, to not hear them" (Brenner 1978, p. 135). Hence, even when interviewers know 
that respondents misinterpret questions or give inadequate answers, the rules in a standardized 
interview may be seen as a normative prescription for subordination under respondents' task- 
inconsistent behavior. 
On the other hand, survey practitioners are all too familiar with the frustration with 
interviewers who for no discernible reason skip questions or distort question content beyond recogni- 
tion. In the final analysis, interviewing and interviewer training is a walk on a knife's edge between 
allowing beneficial' adaptations of the standardized questionnaire to the 'unstandardized' individuals 
and avoiding bias-generating distortions. While in this paper I cannot even begin to systematically 
evaluate social interaction in the interview, the nature of my data makes it possible to construct crude 
those with realistic expectations about the sample distribution would then win the prize. 
'Most radically in the self-administered CASI (computer-assisted self-administered 

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.

powered by Goobi viewer