When
Leveraging the Reinforcement Framework to Improve the Quality of Classroom Communication in Mathematics: Treatment effects from a 3-Year Intervention Study
Abstract: In this study, we operationalized the Reinforcement Framework for an intervention to improve the quality of classroom communication in mathematics. Eighty-four second and fourth grade elementary school teachers from across the US were randomly assigned to treatment and control conditions after baseline. The intervention goal was to put teachers back into “learning mode” by engaging in small daily explorations during an introductory classroom activity, thereby gaining new information on which to reflect and learn from. To promote this reflection and increase awareness of their own internal models of classroom communication, teachers participated in a total of ten online meetings during the 2 intervention years and submitted weekly feedback logs. These logs detailed one of their explorations, the information gained, and what they learned. During each year of study, teachers recorded three pre-selected introductory activities, which we coded for the mathematical quality of classroom communication, including the specific mathematical ideas teachers and students engaged with. Mixed-effects ANOVAs indicated that intervention teachers’ overall quality of classroom communication improved compared to control teachers. We further found that the quality of mathematics and amount and quality of student contributions improved in intervention classrooms relative to controls. Bayesian networks analyzing the coded mathematical ideas and events showed notable changes in connections for teachers in the treatment group, providing evidence of changes in teacher actions, which may be interpreted as internal value and model updates. Our findings provide preliminary evidence for the applicability of the Reinforcement Framework to model decision-making and learning in real world contexts.