I had the privilege of observing the second-ever live Software Carpentry Instructor Training workshop last week at the University of Virginia; that work continues to refine a solid best-of in educational psychology with which to arm our new instructors. I’ve seen this content several times now, and some themes and patterns are beginning to come together as guiding principles that not only lay the foundations for the SWC instructor pool, but give some direction for our broader skill-building efforts in going forward. Those two themes are feedback and student-driven learning, and they both fall under the umbrella of an adaptive school of teaching.

Much discussion of the importance of feedback for instructors in professional development has come on the heels of Greg’s much-cited blog post of late; the Science Lab’s own Instructor Hangouts is one prong in that effort. The UVa workshop advanced this discussion by helping to operationalize feedback; two strategies were offered:

• Outside-In analysis, organizing criticism into physical, presentation style, and content-oriented comments, to be addressed in that order, the idea being to quickly address some of the more concrete criticisms, before moving on to the more abstract.
• Critique Checklists, simply having a checklist of past mistakes to have an assistant watch for while you present, in order to focus the attention on problem areas and get the critic thinking about what to watch for, rather than leaving the assessment practice completely vague and unspecified.

Both these ideas boil down to an effort to make the feedback process more systematic, and thus more exportable to instructors of widely varying experience; by systematizing feedback, less experience and mastery is needed to provide constructive critique, and drastically reduces the conceptual overhead of the process.

But beyond setting up workflows for generating feedback, the idea that came up several times at UVa and which has captured my attention of late is that of the various forms of student-driven learning. Peer instruction, which at its heart creates micro-tutoring sessions between weaker and stronger students, and in-situ course corrections for lectures guided by formative assessments carefully crafted to expose student misconceptions, are not very new ideas anymore. But when taken in context with student’s dramatic preference for collaborative activity over lecture in class, the need to address a skills range that simple oratory cannot accommodate, and the deft success of the PyLadies study group model (all of which we discussed at Instructor Hangouts recently), I begin to wonder: should squad-based learning become our primary vehicle of instruction? For all the positive studies and real-world mileage we’ve gotten out of these ideas, we are still lecturing at our workshops – even though we know that students are more successful when they come and learn collaboratively in groups of peers, even though our classes are often too varied in practice to cater to a particular discipline or a particular skill level with a single broadcast of information. Is it time to change?

The UVa workshop advertised that a good lecture / exercise cadence is to speak for no more than 15 minutes at a stretch before turning students loose on an activity; that balance has come a long way since my undergraduate days. I think the Mozilla Science Lab is positioned to push this trend even further. Lecture will always have a place in our workshops, if only to display the body of knowledge we want our students to reach for; but as we explore other teaching channels, the inclusivity and adaptivity of small cohorts of peers is emerging as our north star.