A comparative analysis of results on programming exams

Measuring student performance on assessments is increasingly important, especially when mapping outcomes to particular topics in a university subject. In this paper we investigate the classification of exam questions. In particular, we examine the performance of students in two programming subjects, as a means of determining how we can measure the difficulty of a particular question. This can not only serve as a calibration for the expectations of instructors about the difficulty levels, but also as a means of examining what it means for a question to be considered difficult.