Comment of the Day: Tracy Lightcap: A Note on Coursera CEO Rick Levin's Clark Kerr Lecture...: "I think Thoma's course is indicative of the kinds of courses that succeed on-line... http://www.bradford-delong.com/2017/05/highlighted-rick-levin-kerr-lecture.html?cid=6a00e551f08003883401b7c8fbb17e970b#comment-6a00e551f08003883401b7c8fbb17e970b
...There appear to be two varieties: required courses and puzzle courses. If an on-line course is required, then students will keep with it in hopes of fulfilling the requirement. Since most of theses courses are entry-level, they are mainly concentrated on delivering facts and fitting them into descriptions. This is admirably suited to on-line work.
The second kind of course is one that is concentrated on puzzles. Most introductory math and computer science courses are like that. Students are introduced to a variety of techniques that can be used to solve puzzles that arise in a particular field. There's good evidence that this works well.
What doesn't work very well are courses that teach how to apply the techniques found in puzzles to analytical reasoning. This is one reason why you don't hear much about success in even basic lab science courses on-line. Face-to-face courses work much better for such advanced work. Such courses can work with a "hybrid" model too: two days looking at vids and on-line material + one day of in class work. Problem = this doesn't improve efficiency much at all.
I've read a good deal about this and there doesn't seem to be an answer. Purely on-line curriculums seem to produce both high drop out rates and poor evaluative results. If nothing else will do and the course of study has a lot of puzzle oriented courses (computer science seems to fit the bill) then success can be closer to face-to-face and hybrid courses. But the gap never seems to close. This may be one reason why the rate of increase in on-line students is leveling off, especially in the two-year colleges that it would seem to be most suited to...