From Teaching Open Source

This was my research statement submission for the PhD program in Engineering Education at Purdue. I was admitted (I started my studies in Fall 2011) and granted a Purdue Doctoral Fellowship, so something must have gone right.

I'll put it bluntly: I don't have formal research experience of the type most undergraduates applying to graduate school get, and I've never been a co-author on a research paper. My college was so new when I started as part of the 2nd graduating class that we simply didn't have labs to do research in, so I spent my "undergraduate research experiences" sourcing and installing cameras in the design lab, configuring new servers, and turning blank rooms into environments where later students came to do excellent research.

My exposure to the world of academic research, conferences, and publication largely comes from my faculty mentors in the http://teachingopensource.org (TOS) community, a group of open source contributors and university professors who are experimenting together with student involvement in open source projects. I thank these mentors profusely, as they've gone out of their way to welcome me into their homes and classrooms and to coach me through submitting papers and sitting on panels with them at academic conferences. We have several driving questions which we attempt to answer through our actions: what is "the open source way" of doing things that make open source communities (largely software, but also hardware, design, writing, and others) such dynamic environments for learning, what aspects of this methodology can be translated to other fields and applied to undergraduate engineering education in multiple disciplines, and what results follow from these experiments?

Most of my colleagues in the TOS community are academics, so I add my own ending to this question: "and what can open source communities learn from academia?" Good teachers are experts at scaffolding and bringing newcomers into an unfamiliar domain, skills that open source communities are notoriously poor at. The methodology for newcomers brave enough to venture into our turf is "sink or swim - oh, and our instructions are written for advanced swimmers." Academia has frameworks for reflecting on and evaluating learning that could make "the open source way" more accessible to the many disciplines that stand to benefit from it, and more beneficial to current contributors who may lack guidance on how to formally reflect on the immense amounts of learning they are (largely unconsciously) absorbing every day, and I would like to experiment with transfering these frameworks back into open source communities and seeing whether their success - as measured by those communities themselves - rises as a result.

I've only begun learning how to frame these questions in a researchable way. (Indeed, it was a revelation to me that not all questions were suitable for good research.) A crude first attempt at tracking metrics can be seen at http://teachingopensource.org/index.php/POSSE_stats#The_data, where I compiled follow-up results (from publicly available data) on alumni of the faculty workshop on open source community participation I've taught for the past 2 years. A good first research project might be surveying NSF grants given to groups to run faculty workshops and replicating their methods to study the effectiveness of the workshops I've taught; however, I'm more than open to different starting points.