A web-tool to build project teams based on student's preferences

In 2014, I developed a program based on Genetic Algorithm (GA) to help instructors teaching design classes assigning students to sponsored projects. For the last 5 years, this program has been successfully used by three instructors, helping more than 1000 students, increasing the overall student satisfaction in the class, and reducing the number of hours spent by the instructor creating the teams (from 1-2 days to a couple of hours). Students are no longer assigned to projects that they did not select, and about 80% of the students are assigned to projects they had given a high weight. In surveys, students report they are highly satisfied with their team and project assignments at both the beginning of the semester and after the semester. Indeed, the distribution of student satisfaction with their team assignment becomes even more positive upon completion of the course. Students also indicate informally that they appreciate being able to select one classmate to be part of their team. In Summer 2019, I started working with a colleague from Computer Science and a couple of undergraduate students to create an open-source web-based tool for this algorithm that can be used by others. We have already instructors from two engineering departments interested in using the web-tool in Spring 2020.

Mohan A., Dey P., Tan S., Johnson B., Fagen-Ulmschneider W., Silva, M., “Introducing junto: a web tool to build project teams based on a bidding strategy”, 127th American Society for Engineering Education Annual Conference and Exposition (ASEE 2020), 2020.

Contact information

Mariana Silva
2213 Siebel Center
(217) 300-6633