Dive Brief:
- MIT researchers have come up with a new way to grade programming assignments in large classes.
- The program, OverCode, analyzes students’ programming all at once, grouping the ones that use the same techniques and giving faculty a chance to see where the differences are across groups, Campus Technology reports.
- In one study, Overcode allowed reviewers to get through 45% of student responses to the most difficult of three assignments while their peers reviewing assignments one at a time covered just 9%.
Dive Insight:
OverCode has the potential to increase faculty feedback for students in large classes. As Campus Technology reports, it is easy to assign coding homework and then review student programming for the right result. It is significantly harder to parse a student’s code for the sections in which they went wrong — especially if there are hundreds or thousands of students taking a course. OverCode isn’t ready to market yet, but the results to initial usability studies are promising.