This study measures the effect of a model of explicit instruction (EMIT) on the: 1) physics graduate teaching assistants’ adherence to reformed teaching methods, 2) impact of the instructional model on GTAs’ beliefs about the nature of physics and physics problem solving and 3) undergraduate physics students’ understanding and performance in an introductory calculus-based physics course. Methods included explicit modeling for the treatment group GTAs of the Reformed Teaching Observation Protocol (RTOP) and assessment of treatment and control GTAs and their students throughout the semester. Students’ understanding was measured using the Force Concept Inventory (FCI) and Flash-mediated Force and Motion Concept Inventory (FM2CA). Students were surveyed about performance of GTAs using the Student Survey (SS). Results indicated changes were tied to individual GTAs’ beliefs about the nature of physics. Student conceptual understanding reflected a two-fold Hake gain compared to the control group. General application of the EMIT model presupposes explicit instruction of the model for GTAs.

This study measures the effect of a model of explicit instruction (EMIT) on the: 1) physics graduate teaching assistants’ adherence to reformed teaching methods, 2) impact of the instructional model on GTAs’ beliefs about the nature of physics and physics problem solving and 3) undergraduate physics students’ understanding and performance in an introductory calculus-based physics course. Methods included explicit modeling for the treatment group GTAs of the Reformed Teaching Observation Protocol (RTOP) and assessment of treatment and control GTAs and their students throughout the semester. Students’ understanding was measured using the Force Concept Inventory (FCI) and Flash-mediated Force and Motion Concept Inventory (FM2CA). Students were surveyed about performance of GTAs using the Student Survey (SS). Results indicated changes were tied to individual GTAs’ beliefs about the nature of physics. Student conceptual understanding reflected a two-fold Hake gain compared to the control group. General application of the EMIT model presupposes explicit instruction of the model for GTAs.