Method to study student engagement uniformity in a classroom

We present a method to study engagement level uniformity in a class of students. We validate our method by comparing two semesters taught using different methods in a physics and mathematics course. The first semester used conventional methods while the second semester used graphic spreadsheets-based laboratories and homework before the introduction of formal methods. We found that engagement of the students that were below the median at the beginning of the semester improved by an order of magnitude (by our metric) while the engagement level of students initially above the median did not change.