Distributed expertise is an instructional approach in which each student specializes in one topic area that falls within the broader content goals of the course. Students spend a portion of the course gaining expertise in their focused topic area, and then join together with one student from every area to form a mixed-expertise group for a major culminating project. This final project is designed to require knowledge developed in each of the specializations, thus leveraging and requiring the expertise distributed throughout the group.
This is distinctly different from a traditional content delivery model where the goal is to teach all of the students the same content throughout the duration of the course. In a distributed expertise model, the goal is to enable all students to develop the same fundamental skills and content awareness, but allow them to specialize and gain deeper expertise in just one content application area. Distributed expertise closely emulates the way sports teams and many industry collaborations occur. In fields outside engineering, distributed expertise and jigsaw learning are used to promote authentic student discussions and increase active learning. However, these techniques have not yet been widely applied to teach computational thinking.
My current research into distributed expertise focuses on an analysis of four sections of the Introduction to Computing in Engineering course, which is required of first-year engineering students at a research university in Massachusetts. The overall objective of the course is to teach students how to apply computational tools to engineering problems and tasks. The course was taught by four different professors. One professor used a distributed expertise model and the other three used traditional content delivery methods. In the distributed expertise course, students were broken up into four specialty groups: computer vision and image processing, sensing and actuating, data acquisition and processing, and system integration and communication. Students spent the beginning of the semester learning as one large group and then broke into their specialty groups for the middle portion of the semester. They then spent the remainder of the semester working in final project groups that consisted of one student from each of the specialty groups.
Data analysis is ongoing, but early results show that using a distributed expertise model was not only effective at teaching computational thinking skills, but also that it improved group dynamics and led to high solution diversity. Check out the Work in Progress Paper presented at the 2019 First Year Engineering Experience Conference (FYEE) for more information on the study and early results.