Abstract
An attempt is made to point out and clear up some of the major conceptual and methodological issues in the study of role‐taking accuracy. Three conceptual issues are discussed: the relationship between roles and role‐taking, the distinction between role‐taking and related concepts such as empathy and sympathy, and the distinction between role‐taking accuracy and role‐taking ability. Methodological criticisms by Cronbach and others of the type of measure that is typically used to measure role‐taking accuracy are reviewed and evaluated. It is concluded that these criticisms pertain mainly to the measurement of role‐taking ability and empathy, and not to the measurement of role‐taking accuracy.