Eberly Center

Evaluating The Use Of Transactive Exchange To Form Effective Teams In A
Computer Science Course

We have designed an experiment to evaluate the efficacy of using collaborative
exchange as a mechanism to suggest effective teams in a computer science projectbased
course.
Collaborative exchanges are used to calculate transactivity - the reasoning of one
utterance that operates on the reasoning of another [1], which is known to be linked to
effective learning in collaborative groups as it captures learners’ building off the
contributions of their learning partners [1, 2, 3].
In the cloud computing course, students complete ten individual projects and a 7-week
team project. At the end of each individual project, each student authors a short
reflection piece which other students then provide feedback on serving as a collaborative
exchange.
A constraint satisfaction algorithm [4] globally maximizes the transactivity that was
observed in these exchanges to suggest team members. The effort is to evaluate
whether the suggested teams perform better than self-formed teams.
[1] Berkowitz, Marvin W., and John C. Gibbs. "Measuring the developmental features of
moral discussion." Merrill-Palmer Quarterly (1982-) (1983): 399-410.
[2] Teasley, Stephanie D. "Talking about reasoning: How important is the peer in peer
collaboration?." Discourse, tools and reasoning. Springer Berlin Heidelberg, 1997. 361-
384.
[3] Wen, Miaomiao, et al. "Transactivity as a Predictor of Future Collaborative
Knowledge Integration in Team-Based Learning in Online Courses." EDM. 2016.
[4] Wen, Miaomiao. Investigating virtual teams in massive open online courses:
deliberation-based virtual team formation, discussion mining and support. Diss. Carnegie
Mellon University, 2015.

Sankaranarayanan, Sreecharan
Rosé, Carolyn
SCS, LTI
Dashti, Cameron
Sakr, Majd
SCS, CSD