Personalizing Math Instruction At Scale: A randomized controlled trial of MATHstream

Declines in math learning present a significant challenge to the future well-being of students, exacerbating educational and professional disparities among low-income students and their peers. We aim to support struggling middle school students by evaluating the effectiveness of MATHstream, an online active learning solution developed by Carnegie Learning. This high-quality and cost-effective EdTech solution focuses on closing math knowledge gaps and promoting equitable student outcomes. Our study will contribute valuable evidence to public policy efforts by assessing MATHstream’s capacity to enhance math learning at scale. Our primary research question is: To what extent does access to MATHstream improve student math outcomes? We hypothesize that students using MATHstream will outperform their peers in a district-administered, standardized end-of-year math assessment. To test this, we will employ an experimental design with randomization at the teacher level and stratification by school. We intend to conduct this study in the 2026-27 school year. We aim to recruit 30-35 schools (240-280 teachers), leveraging Carnegie Learning’s existing school partnerships.

This project is supported through the Learning Engineering Virtual Institute (LEVI), in collaboration with J-PAL North America, which builds off of a previous pilot project.