SRI Authors: Nonye M. Alozie
Alozie, N., Dhamija, S., McBride, E., & Tamrakar, A. (2020, June). Automated collaboration assessment using behavioral analytics. In review. International Conference of the Learning Sciences (ICLS) 2020, 2, 1071-1078. International Society of the Learning Sciences.
Abstract
The 21st century skills and STEM learning standards include collaboration as a necessary learning skill in K-12 science education. To support the development of collaboration skills among students, it is important to assess and support students’ proficiency in collaboration. We present the process of developing a tool that assesses collaboration quality based on behavioral communication at individual and group levels. The assessment tool uses behavior analytics comprised of multistage machine learning models built on an intricate collaboration conceptual model and coding scheme. Our collaboration conceptual model shows how layers of behavioral cues contribute to collaboration and serves as the foundation of an automated assessment tool for collaboration. We present initial findings that show reliability between our assessment of behavioral interactions with and without speech. An automated collaboration assessment tool will give teachers information about student collaboration and help inform instruction that will guide and support students’ collaboration skill development.
