Knowledge Bank

The analyses presented here deal with the interviewer's effect in survey research. An attempt is made to estimate its size basing on the data from the Polish General Social Survey and information on selected individual characteristics of interviewers. I took into account their sex, age, marital status, education, income, occupation and work experience in interviewing. On the one hand, any significant correlation occurred between interviewers' characteristics and certain selected variables on the part of data collected by them. However, on the other hand, interviewers' characteristics - as token all in all - account for a significant proportion of the total variance of the criterion variables. It amounted to 11-27% for questions on 'objective facts' and 39-63% as concerns attitudinal questions. In discussion of the results the caution is made that so high interviewer's effect may be inflated by some aspects of situation of interviewing that could not have been token controlled for.