Automated testing of programs

To determine the effectiveness of our instructional processes, we will archive all scoring and feedback on student work at the finest level of granularity, i.e., at the problems/question level. For reasons of cost, timeliness, uniformity and accuracy, it is important that, insofar as practical, we automate the grading of both proctored and unproctored student work. It does, however, place a burden on those assigning student work to make reasonable scoring automatable.

Scripts to automate the testing of programs are now under construction. To facilitate accurate scoring it is almost essential that grading scripts be made available to students prior to assignment deadlines to use with their own test cases.