University of Michigan School of Information
Scalable Remote Peer Help for Programming Education
Group office hours where students work together on problem sets offer great opportunities for collaborative learning. Students avoid the frustration of getting stuck for long periods of time, and students learn by explaining things to each other. This project aims to bring some of those benefits to students in introductory programming courses without requiring co-location. This will be especially useful for course offering at colleges with many commuter students who would find it harder to take advantage of group office hours, and to global learners in online courses.
We propose to extend chat.codes, a tool that facilitates remote conversations about code, and integrate it into the Runestone interactive textbook environment. Students will see help requests associated with particular exercises when they are working on those exercises. Once they click through to the conversation about a particular help request, the conversation will be indexed to code runs resulting from an editor shared by all participants. In addition, we will develop a tool that coaches students on how to formulate questions well. In a second stage of work, we will expand peer help beyond on-campus cohorts. To take advantage of the larger pool of potential helpers, we will match help requests more selectively with helpers.
The amount of the award is $598,926 for the project period. The grant is funded by the National Science Foundation.