#502: A “Co-Design Intervention” with AI analytics to Support the Academically Productive Talk of Title 1 School Math Tutors and their Students: A Comparative Interrupted Time Series Analysis

This study examines whether a co-design-based professional learning intervention for tutors, supported by an AI feedback tool, influenced students’ Academically Productive Talk (APT) within a high-dosage algebra tutoring program in Title 1 schools in an urban U.S. school district. Using a comparative interrupted time series (CITS) design, we analyzed 9,037 tutoring sessions over the 42 weeks of one academic year, comparing tutors who participated in a nine-week co-design process with those who did not. Tutors in the intervention co-designed an AI interface called the “Talk Trees”, visualizing student talk patterns to guide student reflection on collaborative discourse. Results show short-term, positive shifts in students’ relating to academic ideas of their peers following the intervention, with modest, non-significant gains in instances of students making claims and providing reasoning. These findings highlight co-design’s potential to foster dialogic pedagogy in the context of boundary-crossing professional learning. They also demonstrate how CITS methods can evaluate AI-mediated interventions in authentic educational contexts.