Research questions on MOOCs: Who’s talking, who’s completing, and where’s the teaching?

In the last three weeks, I was asked several times at MIT and Stanford about what questions I would like answered about MOOCs.  I didn’t get any answers, but folks at Georgia Tech were asking me about the questions, so I thought I’d share some of them here.  This is the evidence I’m looking for.

What’s the value of the discussion forum in a MOOC?  What percentage of students participate in the discussion forum?  What’s the correlation between participation in the forum and completing? I’ve heard people say that that’s where the “teaching” and learning takes place, but I would like evidence of that. I’ve mentioned here that I used to work in computer-supported collaborative learning in the early 2000’s, and some of our work showed that students posted about 1 note every two weeks most on-line forums (even when students were measurably learning). That’s not really enough posting for a dialog and collaborative learning. Is it the same for MOOCs?  We also found that students were LESS likely to participate in on-line collaborative learning forums when the subject was Engineering or CS.  Is that true for MOOCs as well?

What are the demographics of completers in a MOOC?  We already have a problem of CS being mostly White/Asian and Male.  (Lots of reasons why that is a problem: From equity and fairness, to tapping into the fastest growing demographics vs the fastest shrinking demographics, to involving more diversity in design decisions.)  I suspect that MOOCs are even more Male than face-to-face classes (based on the factors that we know lead to broadening participation in computing).   If I’m right, then investing more in MOOCs is counter to our community’s goals to broaden and diversify computing.

Is there any teaching going on?  Overall, I’d really like to know more about the characteristics of people who complete MOOCs, e.g., how many are working full-time when taking the MOOC, how many hours a week are spent on homework in order to complete, what is the background of completers in terms of other degrees?  Right now, MOOCs are just for autodidacts.  Do we want Computer Science to only be for autodidacts?  Don’t we believe that teaching allows people to succeed who might not succeed on their own?  (Isn’t that the definition of scaffolding?)  Are MOOCs really teaching, or are they filtering out the people who couldn’t learn on their own?  If we want MOOCs to be for more than just those who don’t really need the teacher anyway, then we need to measure who is going in, what they know already, and what they learn at the end (and if they “come out,” i.e., complete).

A new resource for CS teachers doing Peer Instruction

I’m a fan of Peer Instruction.  I’m sharing this announcement that Beth Simon just made on the SIGCSE mailing list about a new resource for CS teachers who want to use Peer Instruction:

http://peerinstruction4cs.org

This website serves to support computing instructors implementing Peer
Instruction — a very specifically designed pedagogy developed by
Harvard physicist Eric Mazur (read more under “About”).  In findings
to be presented at SIGCSE 2013, we report on Peer Instruction’s impact
in reducing class fail rates by more than half and present results
from a quasi-experimental study where students in a course adopting
Peer Instruction scored 5.7% better on the final exam than a control
section using standard lecture approaches.

We hope you might find these resources helpful and discuss them with
your colleagues.  In particular: If you are interested in participating in an e-support
program for faculty adopting PI, we encourage you to sign up on our
web site.   Not only can you get feedback from experienced PI
instructors, but you can also share things that worked with others and
complain about things that didn’t work!