Predicting Others' Knowledge: Judgment Conditions Affect the Accuracy of Estimates of Difficulty for Others

• In Event: 42.083 - Roundtable Session 17
  In Roundtable Session: 42.083-13 - Models for Cognitively Diagnostic Assessments

Abstract

Teachers’ abilities to accurately predict the difficulty of content for students is a crucial component of effective teaching. I examined how the accuracy of those judgments are affected by the conditions under they are made. Subjects estimated the difficulty of trivia questions for others. Some subjects were required to answer the questions, and some subjects were provided the correct answer. Providing subjects with the correct answer improved the accuracy of judgments, but the impact of requiring subjects to answer the questions depended upon the characteristics of the question. Predictions of others’ knowledge were dynamically generated by grounding estimates of others’ knowledge in outcomes of metacognitive processes directed towards oneself.

Author

• Jonathan Garrett Tullis, University of Arizona