BAMOS Vol 31 No.2 June 2018 | Page 23

BAMOS Jun 2018 Research corner with Damien Irving Collaborative lesson development If you’re a regular reader of this column, you’ve no doubt heard me talk about The Carpentries—a global community committed to teaching foundational computational and data science skills to researchers. Hundreds of Software Carpentry and Data Carpentry workshops are held around the world every year, which is a monumental effort for an initiative that only got started in earnest about five years ago. While these workshops have had a massive impact on the computational literacy of the research community, in my opinion the most revolutionary thing about The Carpentries is not what we teach, but how we teach it. More specifically, the revolution lies in what we do in the background to develop and maintain our lessons. When a volunteer instructor is preparing to teach a workshop, they have an extensive collection of open and easily accessible lesson materials to select (and edit/adapt) from. Unlike a static textbook produced by a small group of authors, these lessons are continually refined and updated by a large and diverse community of contributors, which means they are (by a wide margin) the state-of-the-art lessons in their discipline. This process of community lesson development is probably best explained by reflecting on my own personal experience. I participated in The Carpentries instructor training program back in 2013, which among other things gave me a grounding in various evidence-based best-practices of teaching (i.e. an understanding of the fundamental pedagogical principles underpinning the lessons). Upon teaching my first few workshops, I started to contribute back to the lessons by fixing typos and other minor issues identified by participants. As I got more experience with teaching the materials, I started to make more substantive contributions, proposing and participating in re-writes of entire sections. Now with over twenty workshops under my belt, I’m writing a whole new set of Data Carpentry lessons specifically for atmosphere and ocean scientists (see here for a sneak peek). The contrast between this process and that typically followed by university lecturers could not be starker. While contributing back to the The Carpentries lesson materials can be tedious at times, it is much less work (and results in a far superior product) than if I had to develop the materials for my workshops myself. Most lecturers get some hand-me-down materials from the staff member that went before them (if they’re lucky), and then they’re on their own. What’s more, anything they learn about teaching their discipline better has little impact beyond the four walls of their own classroom. By following a community-based approach to lesson development similar to that pioneered by The Carpentries, it’s not hard to imagine that the quality of teaching at universit ies could be improved, while at the same time saving substantial time for lecturers. For instance, I can think of at least five universities that teach detailed courses on the weather and climate of Australia. It would be great to see the teachers of these courses work together (perhaps facilitated by AMOS) to collaboratively develop and maintain a set of lesson materials. Of course, one critical topic I’ve skipped over here is all the ingredients required to make community lesson development work. What platforms are best for hosting the materials? What do you do when people disagree on the direction of the lessons? How do you structure the lessons for ease of use and contribution? To try and assist people with this, a bunch of Carpentries people (myself included) got together recently and published Ten Simple Rules for Collaborative Lesson Development (Devenyi et al., 2018). It distills everything we’ve learned over the years and is hopefully a useful resource for anyone thinking of giving it a try. Devenyi GA, Emonet R, Harris RM, Hertweck KL, Irving D, Milligan I, Wilson G (2018). Ten simple rules for collaborative lesson development. PLoS Computational Biology. 14(3), e1005963. doi:10.1371/journal.pcbi.1005963 23