Response effects in polite cultures - A test of acquiescence in Kazakhstan

Some individuals may be predisposed to agree or acquiesce more than others. If the predisposition is cultural, then studies of public attitudes that rely on questions with agree-disagree response sets may mistake response effects for substantive differences among ethnic groups. In this study, I report the results of six experiments in question form conducted on a 1997 nationwide survey of 1,986 adult (age 18+) Kazakhstanis, 47 percent of whom are Kazakh and 34 percent of whom are Russian. Acquiescence bias is found among the entire sample, but it is stronger for ethnic Kazakhs than for ethnic Russians. Acquiescence bias is thus a problem of both question format and individual proclivities. Attitude statements with agree-disagree response sets are less valid measures of public attitudes than balanced questions with forced-choice response alternatives, and their use could cause erroneous inferences about ethnic differences in attitudes.