PREDICTING WHAT OTHERS KNOW: TEACHER’S ESTIMATION OF MIDDLE SCHOOL STUDENT KNOWLEDGE

Abstract

In this experiment, I analyzed how accurately teachers could predict the physics knowledge of a general population of middle school students. The purpose of this study is to understand how teachers estimate student knowledge and whether the judgment circumstances change the accuracy of those estimates. Further, I will examine how biases impact the outcome of the predictions. For example, an expert in a certain field may predict that others know more about it than they actually do because the expert knows the information. In this experiment, we will ask science teachers a series of questions about predicting what their students may know about physics. There will be two conditions involved in the experiment. Under one condition, the answer with the question was given to the instructor, which prevents the instructor from answering the question themselves. They then estimated the likelihood of students knowing the correct answer to the question on a 0%-100% score. In the other condition, instructors answered the question for themselves and then saw feedback about the correct answer. They then recorded the likelihood of students knowing the answer to that specific question. Our results showed that manipulating the conditions of the experiment did not impact the resolution the participants displayed. However, the conditions had a marginal effect on calibration and a significant effect on estimation of student knowledge.