Our Research
Our Research
We are working to develop AI Partners that collaborate with teams of students and their teachers to promote development of AI Literacy, STEM competencies, and 21st century learning practices of collaborative problem-solving and critical thinking.
The AI Partners are continuously improved through classroom design and development studies. The goal is that our AI Partners will promote deep conceptual learning via rich socio-collaborative learning experiences for all students.
Our research is organized around three core research strands, all seamlessly integrated through a cross-strand focus on developing specialized AI Partners.
Learn more about our research Strands and themes by following the links below!Ìý
Strand 1 - Technical Foundations:ÌýOur Machine Learning Development Experts focus on the development of computational models to process human language and nonverbal signals (such as eye gaze and gestures) in noisy classrooms. This allows us to develop sophisticated control policies that ensure that our AI Partners interact unobtrusively and productively with students.
Strand 2 - Socio-technical Foundations:ÌýOur Team Science ExpertsÌýstrive to better understand how students and teachers can collaborate effectively with AI in both classrooms and remote learning contexts. We investigate how to properly calibrate trust in AI. This involves building robust frameworks for privacy and ethics, as well as creating transparent "under-the-hood" environments where students can actively explore and experiment with AI reasoning.
Strand 3 - Broadening Participation Through Co-Design:ÌýOur Education Experts work hand in hand with students and teachers to develop AI-enhanced curricula. We focus on directly co-designing ethics-centered AI Literacy curricula with teachers and students. We then study how these materials operate in real classrooms to maximize student agency and critical thinking.
Our cross-strand development of AI Partners serves as the connecting tissue of our institute. Here, foundational and use-inspired research converge to build AI Partners that tackle the most pressing issues that have made collaborative learning so difficult to implement:
- CoBi (Community Builder): A partner focused on relationship-building, visualizing collaborative discourse in privacy-preserving ways to help students talk about and reflect on their classroom interactions.
Ìý - SPARK (Supportive Partner for Agency, Reflection, and Knowledge-building): A real-time scaffolding agent engineered to support small groups during dynamic collaborative activities.
Ìý
The iSAT Design Lifecycle
Effective AI Partners cannot be built in isolated labs; they must be developed within authentic educational environments. Our context-based lifecycle is driven by four key stages of research:
- Discovery: We conduct Learning Futures Workshops and co-design sessions with youth and educators to identify their specific needs, aspirations, and interests.
Ìý - Design-based Research: We iteratively design and refine AI Partners directly embedded in curricula. Through rigorous "stress-testing," we ensure that the overall experience with the AI Partners is intuitive and seamlessly integrates into the daily classroom learning experience.
Ìý - Implementation Research: We conduct longitudinal studies to answerÌýwhat works, for whom, and under what specific conditions.
Ìý - Knowledge Mobilization: We share actionable insights with communities of research, practice, and interest through community-building, workshops, and our Nexus Hub.