Gregor Kiczales is dealing valiantly with problems that arose in his course Introduction to Systematic Program Design on Coursera.
Assessment in a MOOC is extremely difficult. Automatic assessment can check the output of a program but is useless to evaluate design, the topic of this course. The original intention was to use peer assessment for two projects, and a system of rubrics and training was laid out. This was abandoned in favor of self-assessment (!) because people take vacations in August and peer-assessment must be synchronous.
Aside from the usual problems with the forum (lack of civility), this course faced a problem caused by the openness of the MOOC. The course targets true novices but experienced people also take the course. Kiczales had to request: “Please, please, please don’t reply to beginner requests for help with advanced comments.”
