摘要：In this paper we use seven quantitative measures of student performance to compare the performance of students enrolled in three physics courses (two hybrid and one MOOC) that have some common features. We find that, despite the fact that these courses have different audiences, aims, and methods, the measures presented here place the students from all three courses on the same scale and reveal performance similarities. All measures are compared pairwise and the sign of the correlation between each pair is consistent for all courses. The percentage-based measures all positively correlate with each other and with Item Response Theory measure, while the measures based on average number of submissions positively correlate together but anti-correlate with some percent-based and IRT measures. Our findings suggest that for all course types students who get a higher fraction of problems correct tend to use fewer submissions to do so and have a higher IRT skill, while students in a MOOC choose more frequently to not attempt a problem upon opening it than students enrolled in hybrid courses.
