Task-agnostic team competence assessment and metacognitive feedback for transparent project-based learning in data science

Assessing team and individual competencies from team projects' outcomes alone can be pretty subjective. Sharing credit for team efforts equally between team members or differentiating individual contributions based on peer evaluations that might be prone to bias destroys motivation and hinders learning. A fair assessment of individual performances should depend on a formative assessment of a team's process and each individual's contribution to tasks. Such an assessment is time-consuming and only affordable to utilise in small classes. This research serves as a small step to synergise the human and Artificial Intelligence (AI) based educational technology to improve the transparency and effectiveness of collaborative Project-Based Learning (PBL). We introduce a web-bot (BotCaptain) to automate parts of the instructional tasks, present a task-agnostic team competency model, and recommends a set of metacognitive feedback for team members. Study findings have implications for the use of AI in PBL environments.