Charles “Dr Chuck” Severance put together an interesting video on his reflections on Coursera from his “Internet History, Technology, and Security” class. Some areas in the video that I’d call out:
01:42 – 02:45 — Chuck shows us how cozy his instructor’s room is. He expands upon the technology setup here, but it’s basically a couple of HD cameras, a monitor, an electronic pad (Cintiq) on which he can doodle over his slides, and a hodge-podge of software to make it all work. Wires are everywhere, but his desk is surprisingly tidy.
11:30 – 17:00 — we see the engagement curve of the course. As you would expect, there is a precipitous drop from signup to completion of the first quiz, which he uses as a benchmark for “actual signups.” This wasn’t surprising. I routinely sign up for more than I take because there’s no reason not to:
- The welcome video, which is a sort of sales pitch, occurs several months in advance of the actual offering and is not always indicative of how the class will be. For example. As Dr. Chuck notes in his video later (28:55), his was taped before he had “any freaking clue what was going on.”
- The Coursera platform requires that I register to see more information like a syllabus or sample lectures. If a course is a stretch for my background, I may give it a couple of lectures. For example, I lasted about two lectures into the Pharmacology course intended as a med student refresher before no mas. Since there’s no effort to sign up or withdraw, I’m encouraged to — and do! — explore.
- Course offerings slip all the time. I had planned to double down on the delicious Winter 2012 offerings while I had a two week bloc of extra time, but all of them slipped. Once back to work, I had to make up for projects that had waited for my return.
- My schedule slips all the time. I have two teenage daughters involved in a lot of activities. My weeknights are disrupted by playing ballet/hip-hop dad.
What was most interesting about this segment was his (wordplay alert!) course correction. The second week writing assignment didn’t go as planned. He made it optional and subsequently offered extra credit for those who wanted to engage further. 5% of the original pool did.
22:30 – 23:50 – he discusses further the student feedback received on the rubric on the writing assignment and how he handled it. He deserves a lot of credit for being receptive to feedback, making adjustments, and being the “benevolent dictator” in adjusting it enough.
25:45 – 28:50 – Things that worked well. The biggest surprise was how well Twitter was for communicating with him. I agree the multi-take quizzes with auto-morphing questions (and explanations) can be excellent.
29:00 – 34:40 – Things that I would do differently. There was a lot of great stuff in this segment.
- Timeline (under the heading “Faster Lecture Translating”) — I understand his point about wanting this to be a community versus being a self-paced course; however, one of the the biggest problems I’ve had with the Coursera offerings are when the instructor adheres to a rigorous, academic-style rollout of material in one-week buckets. I have the overall amount time to do the work, but IRL (in real life) sometimes beckons and I “lose” a week or weekend, only to have twice as much time the next weekend. I’d suggest as a compromise a rolling, three-week window. Jennifer Widom’s course did this and, I believe, was largely the reason I was able to stick with the rigorous work for the full twelve weeks. An interesting benefit is the folks who could jump on material right away would be able to debug the materials others. This was my experience in the database course.
- Delete threads more often — this would be fantastic. Coursera forums can be very cluttered, especially with so many certificate whining threads. For example:
Nothing inspires wanting to help less than all-caps, whining and lots of exclamation points. The forum software would benefit from some rudimentary user-level customization such as ignoring certain topics (e.g., “Where are you from?” — I do not need a mail every time someone posts to this. The roll-up, however, is great), better threading (because reputation points turns it into a mess as people don’t read the really useful nugget with zero RPs at the bottom), or an RSS-style feed for mobile use. The Computational Finance class is using Piazza (which I swear I keep reading as pizza), which seems to be a little better in this regard.
Forum Owl aggressively moderates over students who spend more time on pursuing a credential than learning. - Extra credit — I think this is a wonderful idea as a structured guide to going deeper into material. I spent a lot of time in Roger Peng’s Data Analysis course tinkering with my own data and got a lot out of the material.
As I was looking through some class materials, I was thinking of further improvements that could be added to the peer review process.
First (and self-serving), some flexibility in the scheduling. A typical assignment starting on day 1 is due on day 7. Peer review begins on day 8 and continues until day 14. The suggestion would be open up the grading early to those who have already turned in their assignments (and agree that they’re “done”) as early as day 4. This would grant some flexibility in scheduling and provide early feedback on the rubric.
Second, allow some commenting so the review is useful to the recipient. In the Data Analysis ][ course, the rubric lacked any provision for an explanation and, as such, was not remedial.
Thirdly, make the rubric available early.
Great info here. I’m following the Coursera experiment with interest myself. I wonder if the various instructors share their lessons learned with each other? I imagine some virtual faculty coffee club where they can swap ideas. I bet this doesn’t happen very often, though, due to schedules…
I’ve entirely stopped reading or participating in the forums, rendering the courses themselves a solo endeavor. The noise level is so high that it’s simply not worth participating. I like your ideas here, but fundamentally I don’t really want to try to follow anything that has thousands of participants — it’s unmanageable. Much better is to have a small study group (or course buddy 🙂 ).