Key points are not available for this paper at this time.
In recent years, Massive Open Online Courses (MOOCs) have become a phenomenon presenting the prospect of free high class education to everybody. They bear a tremendous potential for teaching programming to a large and diverse audience. The typical MOOC components, such as video lectures, reading material, and easily assessable quizzes, however, are not sufficient for proper programming education. To learn programming, participants need an option to work on practical programming exercises and to solve actual programming tasks. It is crucial that the participants receive proper feedback on their work in a timely manner. Without a tool for automated assessment of programming assignments, the teaching teams would be restricted to offer optional ungraded exercises only. The paper at hand sketches scenarios how practical programming exercises could be provided and examines the landscape of potentially helpful tools in this context. Automated assessment has a long record in the history of computer science education. We give an overview of existing tools in this field and also explore the question what can and/or should be assessed.
Staubitz et al. (Tue,) studied this question.