Right now, people all over the world are learning how to write programs and create web sites, but for every one who is doing it in a classroom there are a dozen free-range learners. This group will explore how we, as mentors, can best help them. Topics will include:
What does research tell us about how people learn? Why are the demographics of programming so unbalanced? What best practices in instructional design are relevant to free-range learners? What skills do people need in order to bake their own web? How are grassroots groups trying to teach these things now? What's working and what isn't?
This is very similar to Norvig's classic: Teach Yourself Programming in 10 Years
I read it as do not oversell what you can do, which is different from saying that it's useless. No programming course will not teach you how to be a good programmer, but it can be a start.
Then you can take my course and learn a bit more 
I bristle at that attitude. It's like a "weed-out" attitude that kept me away from traditional courses--and turns away many other students as well. I'd hate to see that attitude translate to online courses. One of the things that online education has the potential to do is to allow for more collaboration--students asking students questions rather than relying on the teacher. While that certainly happens to some extent in traditional courses, it depends, I've found on the culture created by the teacher. Weed-out courses tend to foster a competitive environment, which leads naturally to a lack of collaboration. It would be interesting to see if there's any research on this for traditional classrooms and then also for online courses. I wonder, too, how our free-range students feel about collaboration and competition. And, as Guzdial says, what is the definition of effort.
I agree that competitive environments discourage collaboration, and I find that decreases the quality of the learning experience. However, I've now run a total of 11 three-week sessions with intermediate programmers in which we have dropped students due to inactivity or due to not meeting weekly checkpoints, but for the most part managed to have an *incredible* amount of collaboration between our participants. Every one of our students spends at least a couple hours a week helping the other students, all in a completely open environment where they can overhear one another's conversations. We don't require this collaboration, we only tell them it's allowed and encouraged.
What makes it so that an environment in which you can expect a third or more of the folks who start the course to be "weeded out" by the end of the course NOT become a competitive pressure cooker? I think it's the structure that does it.
Here are some the things we do that I think allow us to make cuts without hurting our learning environment:
- We give each student a LOT of support. We have several mentors and two instructors in each session, so that the Student::Mentor ratio is basically 3:1 or less. Each student can expect to have at least a couple hours a week of one on one collaboration with a mentor/instructor.
- We allow the students to propose their own assignments based on a theme, which means that they can choose something they at least *think* will be successful.
- We don't do one pass over assignments. Students send in their work and then fellow students and mentors review it and make suggestions, then they resubmit. Then instructors look it over, give further feedback and instructions, and then they resubmit again. We repeat the process as much as necessary to get the student to pass the assignment with sufficient quality.
- We provide all of our exercises and lay out all of the requirements for the course up front, preventing the student from running into nasty surprises mid-session that would knock them out.
- We release all of our materials so that prospective students can review them to see if the course would be a good fit for them.
- We carefully design exercises so that they can be worked on in the open without any spoilers, which means that unless they're physically doing the work for another student, it's impossible for our students to "cheat" by helping one another. In fact, collaboration around the dependencies and common issues between their projects is actively encouraged.
These things combined have made it possible to make a course that says "Yes, we do fail people who don't meet the expectations", but at the same time says "If you feel like this is something you'll be capable of doing, you'll have endless support from us to make it happen"
The trouble with this model, and the main reason why I'm taking this course, is because I fear it'll break down hard with novices and advanced beginners. There is a wide space between a simple study group and what we're currently doing, and I don't think this sort of environment we created would be suitable for new programmers without sufficient tweaking.
I want to be able to find a way to make sure we're spending our very limited resources wisely on the students who are "willing to put in the effort" without shutting out those who are trying in earnest but not being well served by our teaching style.
I didn't read that attitude at all, I don't know if it was intended.
I'll note though that I've had courses where collaboration and competition were both present so I don't see them as opposite. In fact, it's a fact of humankind that we are often most collaborative when it comes to beating the other teams (sports being the perfect example). I've had many courses where each group wanted to be better than other groups, but inside the group there was collaboration.
