Automated generation and grading of programming assignments (abstract only)
Steven C. Shaffer, Cliff Shaffer · 2013
A problem with teaching large classes of introductory programming students is that students need copious practice but instructors do not have time to grade thousands of student programs. Large classes might require the instructor to restrict the graded programs to a handful. However, students need much more practice than this, and experience shows that if an assignment is not being graded then many students will not do it. Also, if the same problem is assigned to all students, copying of answers will occur. Programs that generate unique assignments and automatically grade submissions have been developed. Beginning with a brief example, this session will focus on experiences attendees have had with such approaches, and what an idealized system might look like, including exemplar use cases.