Understanding the Interconnected Drivers of Mathematics Test Performance: A Longitudinal Study

Mathematics performance remains a critical indicator of educational equity and academic preparedness in an increasingly data-driven world. This study examined factors affecting math achievement in New York City from 2013-2023, analyzing 3,395,816 student records from Grades 3-8 using NYC OpenData. Clustering analysis revealed two distinct performance trajectories, with higher-scoring clusters linked to periods of targeted intervention and policy reform, while declines corresponded with structural disadvantage and post-pandemic disruptions. Outlier analysis showed substantial shifts in performance levels following grading policy changes, underscoring the impact of systemic reforms. Results revealed persistent achievement gaps of 1.8-2.4 standard deviations across socioeconomic status, disability status, and English Language Learner groups. Ridge regression identified English proficiency, disability status, and economic disadvantage as the strongest predictors of math achievement. These findings highlight the need for long-term, context-sensitive educational strategies that address persistent inequities through both instructional and systemic interventions.