This commentary uses the dynamic of identity-protective cognition to pose a friendly challenge to Jussim (2012). Like other forms of information processing, this one is too readily characterized as a bias. It is no mistake, however, to view identity-protective cognition as generating inaccurate perceptions. The “bounded rationality” paradigm incorrectly equates rationality with forming accurate beliefs. But so does Jussim's critique.

We argue that Henrich et al. do not go far enough in their critique: Sample diversification, while important, will not lead to the detection of generalizable principles. For that it will be necessary to broaden the range of contexts in which data are gathered. We demonstrate the power of contexts to alter results even in the presence of sample diversification.