Further evidence of the relationship between case typicality and diagnostic performance: Implications for medical education