2026 ASEE Annual Conference & Exposition

This is a full empirical research paper for the Educational Research and Methods Division. Collaborative and cooperative learning experiences are a cornerstone of engineering education that offers students opportunities to develop sociotechnical skills necessary for successful professional practices. However, existing research suggests that patterns of inequities in student teams, particularly those experienced by racial/ethnic and gender minoritized students in engineering, frequently threaten equitable participation and learning outcomes in student teams in engineering.

Decades of research have examined pedagogical practices and issues related to instructional design due to the prevailing belief that faculty members are key actors in mitigating teamwork inequities that undermine learning outcomes for historically excluded students in engineering. Scholars have examined the use of team formation strategies, learning analytics in learning management systems, self and peer evaluations, and other approaches to mitigating and remedying inequities in student teams that undermine student learning. However, the literature base frequently notes limitations of such approaches, such as limited faculty bandwidth to monitor all student teams or the tendency for self and peer evaluations to reify rather than remedy inequities.

In response to the shortcomings of these common practices, this research seeks to design social robots for supporting equitable teamwork in engineering learning environments. These social robots are designed to (a) automatically detect when inequities occur in collaborative teams, (b) determine appropriate times and response mechanisms for mitigating patterns of inequity, and (c) respond strategically to promote equitable behaviors. Thus, this paper focuses on human-human data to characterize mechanisms of inequity as a first step to designing and implementing appropriate interventions.

We describe mechanisms of inequity in terms of three categories, with implications for the design and implementation of interventions for supporting student participation and learning. For example, moments in Category 1, passive marginalization, occur when team members exhibit covert behaviors that undermine the contributions of others, and are particularly difficult to detect, even in human-led interventions. Indeed, students may even be unaware of the ways that their behaviors constitute constraints on opportunities for learning for their peers. For example, moments during which students fail to respond to the ideas of their teammates during team discussions, or when other students’ attentional focus is not on the speaker, constitute moments of passive exclusion. Category 2, constrained inclusion, occurs when students are invited to participate in team activities but only to the extent that referent others do not impede full participation. Such moments are frequently covert and can be mistaken for full participation by observers, such as when one’s opportunities to participate are constrained by supervision from teammates. Finally, Category 3, overt exclusion, occurs when student behaviors explicitly exclude or marginalize the contributions of other students. Such moments can occur even in the absence of the subject of the overt exclusion, such as instances of team meetings in the absence of particular team members. In this paper, we suggest instances across these categories require different interventions.