A New Classroom for a New Kind of Computing Student: Brian Defends
Brian Dorn is defending his dissertation this week. For several years now, he has been studying graphic designers who program. He started by studying the information foraging behavior of the graphic designer in the wild. First, he characterized who they were and what kinds of coding they were doing (ICER2006). Next, he studied one of the information sources that they frequented (Adobe’s Photoshop scripting repository) to figure out what kind of informational nourishment they were getting (VL/HCC 2007). He did careful assessment and interviews with graphics designers to figure out what they knew and what they wanted to learn (CHI2010), and most recently (the paper I presented at ICER 2010 last week), he did an interview study to figure out why they won’t enter our classes to get the information they needed.
At this point, Brian knew who his subjects were, what they were looking for, where they were looking, and why they wouldn’t go where he knew the information was. Now, there is a consortium of researchers studying end-user programmers, but for the most part, they’re coming at it from an HCI perspective. How do we make the tools better? Brian wanted to come at it from a learning perspective. How do we make the people better? How do we teach people where they are with what they need? Continuing my (now tired) metaphor, how can he vitamin-fortify (“Now with Vitamin CS!”) the places where they were foraging?
Brian built two different kinds of code repositories. In one, he just had code, just like the repositories that his designers were already using (like the one Adobe hosted). In the other, he provided real cases (based on the design that Mike Clancy and Marcia Linn created). In that one, he included lots of conceptual information about computer science. He wanted to see if his graphics designers would like the cases the same, would be just as effective at writing code, but would also learn something. If adding the CS content made it less pleasant or hurt their productivity, it wouldn’t get used.
He ran everyone through a task, where they had to write some code and answer some conceptual questions, using whatever resources they would normally use. Next, he split the pool into two groups, of roughly equal performance on code and concepts, and gave each group one version of “ScriptABLE” (his tool) — either the code repository form, or the case library form. They did an isomorphic task: About the same complexity, same kind of code to write, same kind of concepts to answer about.
Huge win: Each group liked their resources. No difference in code writing. Statistically significant better learning by the case library-using group.
There are lots of reasons to be excited by this work. First, it’s a study of a seriously non-STEM group of programmers. He has made computing education work with people who have mostly only studied art, with a disdain for computer science. Second, it’s an audience that is much more gender-balanced than most of STEM. Brian now has an approach that works well for increasing the computing knowledge of art-oriented, female professionals who are pretty darn hostile to normal CS classes. That’s quite an accomplishment. Brian’s work is very important for the CS10K effort, because (as I’ve argued previously) on-line learning is critical to achieve that goal.
For our field, it’s a whole new world for computing education. It’s about making things better for computing learning outside the classroom, with people who aren’t CS majors. We mostly look at classrooms, and mostly CS majors. There are many more non-CS majors interested in learning about computing, and most of them won’t enter our classrooms. Brian is showing us a new space for us to work, providing a process for studying our new “students” and new kinds of “classrooms,” and giving us an example of a successful first attempt. Brian has already started his new job as an Assistant Professor at the University of Hartford.