Our first session for this round of Instructor Hangouts was dominated by a topic suggested by Rachel Sanders of PyLadies – how do you make sure a workshop is aimed at just the right level for your students?

The way is fraught with peril. With so little access to effective training in computing, workshops attract everyone from the completely new, to those that have built skills over years of slugging it out on their own, one segmentation fault at a time, and those who may be experienced in other languages but completely green in what our workshops offer; the skills bell curve can be very broad, particularly for larger, classroom-oriented offerings like Software Carpentry, and each of the demographics listed above are looking for something a little different.

SWC has tested and converged upon some effective strategies in the classroom for narrowing the skills bell; by asking students to complete a pre-exam, organizers can group them into beginner and advanced rooms, and run two workshops in parallel; trial and error has indicated that pretesting and not self-selection works best, and that things flow the smoothest when students remain with the same cohort for the whole workshop, and not switch rooms (although Naupaka Zimmerman made a salient point here – what about the students who google the right answer for the pretest, get in the intermediate room, and are forever lost? Perhaps allowing them to drop to the beginner room at lunch on the first day needs to be examined; but I understand Naupaka is already involved in revamping the SWC pretesting procedure, and I trust great things will come from that effort – more on this soon).

But, organizational strategies aside, what do you do when you’re in the middle of it – standing at the front of a class of students who are so lost, they are in grave danger of resigning to remain so forever, or so ahead they’ve completely checked out? I think Rachel’s wisdom from her experience at PyLadies holds the key to the answer.

One of the really cool things about PyLadies, is that many of their offerings sound very collaborative and small-team oriented. Rachel described activities where PyLadies study groups would get together and work on a MOOC, or on coding challenges like Exercism. In their experience, this leveled the playing field for the students, by letting them move along at their own pace, and learn from each other. This was a lightbulb moment for me in thinking about student engagement and addressing the two tails of the skills bell for two reasons.

First, Greg Wilson has ensured my fellow SWC instructors and I are well aware of the concept of peer instruction, and its crucial final step of having students discuss problems in small groups after the instructor explains the correct solution. In such a model, rather than the instructor trying to somehow speak to the very fastest and the most struggling students simultaneously, the tails of the bell curve are turned in on themselves: as the students discuss amongst themselves, the weakest students get personal tutoring from the strongest students, and those stronger students get the invaluable exercise of learning how to make their understanding so clear that it can be verbalized and transmitted, thus benefiting all. I suspect the PyLadies experience is producing much the same synergy of tutor and tutee.

Second, I thought back to my own teaching experience at SWC. To put it bluntly: students hate it when I talk too much. I speak too quickly, go off on weird tangents, and I often get too excited to get to the ‘cool parts’, and overshoot early. Conversely though, they love the exercises and brain-teasers that I give them, and encourage them to chew on in small groups – these exercises are dynamos of motivation and engagement. This reaction from students is not unique to me, either; many instructors have experienced the exact same response. Could it be that the reason the students get so excited by the hands on exercises, is that they get to start benefiting from the proven advantages of collaborative learning, and stop being subjected to a lecture that just doesn’t fit their level? Which begs the question: in a classroom where it is impossible to speak to everyone’s level at once, should we start thinking of peer instruction, and not lecture, as the real primary vehicle to our students’ learning?

Evidence remains to be collected, but experience across many instructors and several different workshop offerings is beginning to align. The sermon model was never meant to educate; perhaps the way forward for teaching at our industrial scale isn’t in massive oration, but in simple, collaborative squads of colleagues.