The first two keynotes of SIGCSE 2010 have had me thinking for the last month about how research about education influences computing education practice. (Michael Wrinn's SIGCSE 2010 keynote on parallelism in computing classes was also thought-provoking and interesting, but in a different way than the first two.) The short answer is that it's not just about transferring knowledge. It feels more like a process of enculturation.
Carl Wieman's keynote (discussed here and here, with slides available) took lessons from science education and showed how they suggest improved practice in how we teach computing. Science educators (especially in physics) have been measuring student achievement, then changing educational practice in order to improve achievement. They can measure just how ineffective simply lecturing at students is (i.e., students only learn about 30% of the new concepts you're throwing at them), and how much more effective are techniques of student engagement like peer learning (i.e., 40-60% of new concepts are learned, and can get as high as 90%). Carl directly related this material to computing educators. For example, he pointed to our practice in introductory classes of ordering the introduction of basic language elements, then simple computing concepts (like arrays), and finally building interesting things. That's backwards. Instead, we should start with a complex and interesting problem, providing libraries, specialized tools and other scaffolding to allow students to succeed at that problem, then introduce the underlying concepts and language elements so that students can achieve the interesting task without all the support. Start from motivating students and using what students already know.
I've been mentioning Carl Wieman regularly in my blog over the last month. One of my colleagues in the Learning Sciences asked me why. Carl didn't invent this stuff. He's not the primary authoritative source for these ideas. I explained that Carl was engaging in an underutilized form of scholarship. In his book Scholarship Reconsidered, Ernest Boyer talked about forms of scholarship beyond "discovery." Boyer talked about scholarship of "integration" which was about making connections across disciplines and scholarship of "application" which has to do with applying research findings to addressing society's needs. Carl was engaging in both of these, by helping computing teachers interpret science education research and by helping us figure out how to apply it to improve our classes.
Sally Fincher's keynote (discussed here and here) took us one step further. Sally addressed the question of why education research often doesn't influence teaching practice. Even when someone like Carl brings the research into our terms, we don't often make the leap to change what we do. Sally explained that education research findings are often "useless truths." Teaching is an embodied, hands-on, concrete activity. We work with students, and we answer their questions, sometimes under pressure and time constraints. Research is about generalization and formalization, stating truths about many classrooms, which end up being about no particular classroom -- and certainly telling me little about my classroom. How do we bridge that gap? Sally suggests the use of story. Narrative knowledge is about providing examples of how the abstract research knowledge has been adapted to work in someone's everyday classroom, which we might recognize as being like our own. Sally gave several examples from her rich research background where people share these kinds of stories with one another.
These two talks really fit together well (though they were presented in the opposite order than I describe them here) to talk about the path from what we know about research to changing how we teach. It's not about discovering new things about learning -- Learning Scientists already know far more about that than most teachers realize. It's about bringing it from the abstract discovery, to integration with our discipline, and through stories, to apply in our classes. Influencing practice is not about discovering something new, but figuring out how that research meshes with my culture, values, and practice.