A recent study of a Coursera MOOC is really interesting in that it implemented a random assignment of student to 2 conditions – one with no teacher interaction with the students and the other with teacher and teacher assistant interaction in forums. The study is

Tomkin, J. H., & Charlevoix, D. (2014). Do professors matter?: using an a/b test to evaluate the impact of instructor involvement on MOOC student outcomes. Paper presented at the Proceedings of the first ACM conference on Learning@ scale conference. Retrieved from http://dl.acm.org/citation.cfm?id=2566245

Via Ana Cristina Pratas, Ulrike Grabe