Acquiescence Response Bias

Acquiescence response bias is the tendency for survey respondents to agree with statements regardless of their content.

Acquiescence response bias could influence any question in which the response options involve confirming a statement, but it may be particularly problematic with agree-disagree questions. Although many guides on writing survey questions recommend avoiding agree-disagree questions, such questions are ubiquitous in survey instruments. An agree-disagree question asks respondents to report whether they agree or disagree with a statement. For example, respondents might be asked whether they agree or disagree with the statement, It is important for the president to be a person of high moral character. Acquiescence response bias is problematic because the interpretation of an “agree” response is very different if respondents are asked whether they agree or disagree with the posited statement than if they are asked whether they agree or disagree with the statement, “It is not important for the president to be a person of high moral character.”

There are a number of explanations for acquiescence response bias. One explanation is that acquiescence response bias occurs partly due to social norms to be polite. Consistent with this, acquiescence response bias is stronger among cultures that put a high value on politeness and deference. Satisficing theory also provides an account for acquiescence response bias. Satisficing theory suggests that although survey researchers hope respondents will answer questions carefully and thoughtfully, respondents may not always be able or motivated to do so. Instead, they may shift their response strategies to minimize effort while providing a satisfactory response to the survey question (known as satisficing). One such strategy involves agreeing with assertions made by the interviewer. Satisficing theory also posits that satisficing is more likely when respondents' ability and motivation is low and when question difficulty is high. Thus, acquiescence response bias is likely to be strongest among respondents low in ability and motivation and for questions that are more difficult, a perspective that is supported by research studying acquiescence response bias.

There are also a number of strategies researchers use to avoid or control for acquiescence response bias. One such strategy is to include multiple items to measure a construct of interest, approximately half of which are worded so that the “agree” response indicates one position and the other half worded so that the “agree” response indicates the opposite position. For example, respondents might be asked whether they agree or disagree with the statement, “It is important for the president to be a person of high moral character,” and then later asked whether they agree or disagree with the statement, “It is not important for the president to be a person of high moral character.” If respondents exhibit acquiescence response bias and agree with both statements, their answers to these two questions cancel each other out.

There are at least three problems with this approach. First, it requires that survey researchers use a large number of redundant questions. This strategy is inefficient and it may be frustrating to respondents. Second, if researchers average responses across questions, this strategy results in “acquiescers” being given scores in the middle of the dimension, and it is not clear that this is appropriate or valid. Finally, as in the case discussed earlier, it sometimes results in respondents being asked whether they agree or disagree with a negative statement (e.g. “It is not important …”). This may be confusing to respondents, as disagreeing with this statement involves a double negative (respondents are reporting that they disagree that it is not important). This is a particular concern because not all languages treat double negatives in the same way, and agree-disagree questions about negative statements may therefore be particularly confusing for respondents for whom English is not their primary language or if questions are translated into other languages.

...