I developed and taught an on-line introductory physics course which is still being taught by colleagues at NCSU. The course is a distance version for in-service high school physics teachers of the Matter & Interactions curriculum created by Ruth Chabay and me. The goal of the distance course is not to train teachers to teach Matter & Interactions in high school (though some are doing that) but rather to give them a contemporary perspective on the intro physics that they teach, whereas the intro physics course they themselves took in college has for decades been taught from a 19th-century perspective (for example, the word "atom" is never mentioned).
In the section "Course for HS teachers" at matterandinteractions.org there is an article describing the course. It is rigorous and challenging. It requires about 10 hours per week of teachers who are already extremely busy. The only reason they can get through the course, given the heavy demands on their time, is that (a) they have had to become very skilled at time management and (b) they are mature learners. Few of the teachers who start the course are unable to complete it but they have to work quite hard. There are weekly deadlines for completing homework. Professional development for teachers often means a couple days of workshops every year, but this two-semester sequence is orders of magnitude more and deeper than that. On the web site you can read reflections on the course by the teachers. Typically, on-line courses have huge dropout rates and huge failure rates. This has made me very dubious about the possibility of on-line courses for unmotivated college freshmen who seem to need the social contract of in-person courses, and the structure imposed on their lives by frequent class periods and tests. However, I was surprised to be told that an on-line intro physics course at Arizona State has worked well, perhaps because regular on-line chat rooms play an important role. When I was in college I took an engineering course by exam after self-study and I took a correspondence course in Italian in preparation for study in Italy (Purdue didn't offer Italian but Indiana University did). These experiences were successful, but I think few of my fellow students would have had the discipline to manage this. I'm now taking my second MOOC, in order to understand exactly what they are and aren't. Ruth Chabay and I together worked through all of David Evans' Udacity course "CS 101". It was absolutely superb. It had a very clear stated goal: "In about seven weeks you will create a small search engine, even if you have never written a program before." Along the way, every new CS concept was introduced and exercised in the context of that clear goal. The presentations were very clear, punctuated frequently with short multiple-choice quizzes. Homework consisted of writing Python functions whose correctness was judged instantly by sending secret input data to the function and checking that the output was correct. By the end of seven weeks we had indeed written a small search engine. I've just finished Part 1 of Eric Haines' Udacity course on computer graphics (Part 2 won't be ready until May 1). It isn't as good at CS 101, but that's an extremely tough act to follow. The format is rather similar, drawing on excellent tools that Udacity provides to its teachers. Homework consists of making rather small modifications to a large JavaScript program that drives WebGL to display a 3D scene in the browser you're working in. Judging the correctness of the homework is apparently done by comparing the pixels produced by your program with the correct pixels. You are shown the difference, with the result that your mistakes are highlighted visually. Very clever, and instantaneous. Part 1 of the course dealt solely with static 3D scenes; I'm curious to see how animations will be judged, a topic to be presented in Part 2. A significant difference between these two MOOCs is that in CS 101 there was real "grading". You collected points for correct homework. In the computer graphics course there is nothing about this. You can keep making homework submissions until you get it right (aided by the clever visual feedback), and no score is shown anywhere. One other connection: Our Georgia Tech colleague Mike Schatz is developing a MOOC for Coursera on intro mechanics with an interesting concept. Students will use their cell phones to capture videos of moving objects which they will then analyze with video motion software (which allows you to extract position vs time information from a video) and model the motion with VPython (vpython.org). The course is loosely inspired by the Matter & Interactions curriculum which Mike has taught at Georgia Tech, in particular the central role of modeling and of computational modeling. So what do I think about all this, having had both teaching and student experience? I think that the most important aspect of MOOCs is that they have stimulated a lot of experimentation and a lot of healthy debate and reflection. I've seen plenty of dismissive comments that they're nothing new, since there have been on-line courses for many years, but I feel that these comments miss some important points. My own on-line course for teachers is by the standards of such courses quite sophisticated, but the muscle that Udacity has been able to muster shows how much farther one can go with real investments in infrastructure, investments unlikely to be made as long as education is a cottage industry. I do remain skeptical about what is possible in the real world with real students, but it's good to try things, because universities do not uniformly provide superb education, especially in large intro courses, yet with very high costs. Bruce
============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
