Stav Atir, Emily Rosenzweig and David Dunning:

Recognizing one’s ignorance is a fundamental skill. We ask whether superior background knowledge or expertise improves the ability to distinguish what one knows from what one does not know, i.e., whether expertise leads to superior meta-knowledge. Supporting this hypothesis, we find that the more a person knows about a topic, the less likely they are to “overclaim” knowledge of nonexistent terms in that topic. Moreover, such expertise protects against overclaiming especially when people are most prone to overclaim – when they view themselves subjectively as experts. We find support for these conclusions in an internal meta-analysis (17 studies), in comparisons of experts and novices in medicine and developmental psychology, and in an experiment manipulating expertise. Finally, we find that more knowledgeable people make knowledge judgments more automatically, which is related to less false familiarity and more accurate recognition. In contrast, their less knowledgeable peers are more likely to deliberate about their knowledge judgments, potentially thinking their way into false familiarity. Whereas feeling like an expert predisposes one to overclaim impossible knowledge, true expertise provides a modest protection against doing so.