An instructor assumes that speaking frequently in seminars is a reliable sign of understanding. Participation counts correlate strongly with an extroversion measure, while quiet students score as well as or better than frequent speakers on essays that require analysis of the same readings. Which evidence most directly contradicts the instructor's assumption?