Abstract
Attitude surveys often use sets of items with identical response scales (e.g., in Likert format) in order to formulate attitude constructs. There is considerable evidence that such a response format can be susceptible to acquiescence bias (i.e., the tendency to agree with survey questions). The specification of a style factor (acquiescence) in measurement models can result in well-fitted factorial invariant models for cross-cultural surveys. Attempts to model acquiescence are confronted with the phenomenon that models with a positive and a negative factor appear to be as likely as models with a bipolar factor and a style factor. We will evaluate the two competitive measurement models, applied to a balanced set of ethnocentrism items from the 1999 Religious and Moral Pluralism (RAMP) dataset. It is argued that the competing measurement models should be evaluated in the context of a theoretically meaningful nomological network.