Where Low and High Inference Data Converge: Validation of CLASS Assessment of Mathematics Instruction Using Mobile Eye Tracking with Expert and Novice Teachers

Classroom observation research and research on teacher expertise are similar in their reliance on observational data with high-inference procedure to assess the quality of instruction. Expertise research usually uses low-inference measures like eye tracking to identify qualitative difference between expert and novice behaviors and cognition. In this study, we used mobile eye-tracking technology to create a low inference quality indicator for the comparison of experienced and student teachers. The distribution of visual fixations on students was measured using Gini coefficients based on the observation of van den Bogert, van Bruggen, Kostons, and Jochems (Teacher and Teacher Education, 37, 208–216, 2014) that expert teachers show better classroom monitoring. Results confirm that student teachers have a higher Gini coefficient than experienced teachers indicating weaker classroom monitoring. However, the Gini coefficient did not correlate in the predicted way with trained observer coding of video footage of the same classrooms using the Classroom Assessment Scoring System (CLASS) (Pianta, Hamre, Haynes, Mintz, & La Paro, 2007) although the mean differences in behavioral management were higher for the experienced teachers as expected. The CLASS dimension Quality of Feedback was significantly related to the Gini coefficient as an interaction with expertise: Only for novice teachers that a high quality of feedback was negatively associated with monitoring of the classroom.