Co-Designing a Teacher Dashboard to Support Evidence-Based AI Instruction During Problem-Based Learning in Middle School Science
Event Type
Late Breaking Results
Health Care
Human AI Robot Teaming (HART)
Surface Transportation
TimeThursday, October 13th8:45am - 9:00am EDT
DescriptionTeacher engagement in students’ learning and problem solving during technology-enhanced, problem-based learning (PBL) has been shown to support student engagement in and learning of difficult science, computing, and engineering concepts and practices. This is important as PBL incorporates real-world, inquiry-based problem solving designed to prepare students for our technology-enhanced workforce. However, there are challenges to adopting PBL, since most student work is observed through mouse clicks and computer interactions that are difficult to interpret and respond to. Human factors approaches can be used to develop artificial intelligence (AI)-driven tools that provide educators with useful information about their students' performance.

We report a detailed case study on the multi-step co-design and development of a teacher dashboard to support and prepare teachers for the implementation of a technology-enhanced, PBL curriculum in their middle school science classrooms. This work presents a novel, end-to-end demonstration of how to engage teachers in the development of AI-driven learning analytics systems that support PBL curriculum approaches in K-12. In the process, we adapt established participatory design techniques and demonstrate new kinds of prototyping methods to address the unique challenges of co-designing learning analytics systems for PBL curriculum applications. Our contributions include the teacher dashboard tool, a description of how the novel tool features were designed and developed throughout the co-design sessions, and a reflection of prototyping methods used to elicit meaningful teacher feedback on the creation of the tool.