EN
Acquiescence is defined as the tendency of respondents to shift their answer towards agreeing rather than disagreeing with items, regardless of the items’ content. As a response style, it can be a significant factor in distorting research results. This article examines the nature, predictors, and consequences of this response style and the possibility of identifying and controlling it. The results of time stability and domain/method specificity of acquiescence suggest that it is more a participant-related than situational-related construct. Two categories of correlates or predictors of acquiescence were identified: a high level of social deference/conformity and a low level of cognitive processing on the part of respondents. The two most common approaches used to estimate acquiescence are the manifest approach, which involves computing the deviation from the median of the response scale, and the latent approach, which uses confirmatory factor analysis with a bifactor model. The advantage of the manifest approach is its technical simplicity when certain criteria are met, but its ability to cleanse the results of the effect of acquiescence is rather limited. The more technically demanding latent approach, by contrast, is useful for its ability not only to identify acquiescence bias but also to reduce bias. In a model analysis using the latent approach, the results were cleansed of the effect of acquiescence.