Uncovering The Impact of COVID-19 Disruptions on StudentsMathematics Achievement: A CART Analysis of Selected PISA 2022 Data

The COVID-19 pandemic disrupted education worldwide, with mathematics learning ‎particularly affected due to its reliance on cumulative knowledge and structured instruction. This ‎study investigates the influence of socioeconomic background and pandemic-related disruptions ‎on mathematics achievement across countries using data from the Program for International ‎Student Assessment (PISA) 2022 COVID-19 module. The dataset included 109,097 secondary ‎students from 17 participating countries, with mathematics performance measured as the ‎average of ten plausible values. Predictor variables included socioeconomic status, emotional ‎impact, perceived learning loss, family support, and access to digital resources. Multiple linear ‎regression analysis was applied to identify independent contributions of each predictor, while ‎Classification and Regression Tree (CART) modeling captured non-linear interactions and ‎threshold effects. Results showed that socioeconomic status was the strongest positive factor, ‎followed by digital access as a modest contributor, whereas perceived learning loss and ‎emotional impact emerged as strong negative influences. Family support showed limited ‎predictive power when modeled together with other variables. CART analysis further ‎demonstrated that students with high socioeconomic status and low learning loss were most ‎likely to achieve higher mathematics scores, while students with low socioeconomic status were ‎consistently classified as low achievers regardless of other conditions. These findings highlight ‎how COVID-19 amplified pre-existing inequalities in mathematics education, revealing that ‎disadvantage and disruption interact to magnify vulnerability. The study underscores the need ‎for equity-focused recovery policies that address both structural socioeconomic gaps and ‎targeted interventions for learning recovery in mathematics‎.