Socioeconomic Disparities and COVID-19 Effects on Student Achievement: Insights from PISA 2018 and 2022

• In Event: Beyond Mere Performance: Educational Equity and Effectiveness in International Large-Scale Assessment Research

Abstract

Objectives. This study aims to investigate the changes in the influence of SES on reading and mathematics competence by focusing on country-level characteristics that may compensate for the effects of crises on educational inequality.
Current research perspectives. The COVID-19 pandemic has significantly impacted academic performance, exacerbating disparities among students from lower socioeconomic status (SES) backgrounds (Betthäuser et al., 2023; Hammerstein et al., 2021). Apart from student-level characteristics, differences on the country-level such as the length of school closures and general economic strength are relevant factors when analyzing the pandemic’s impact on different student groups (Betthäuser et al., 2023; Wisenöcker et al., 2025). Additionally, as digital resources became crucial for learning during the pandemic (Kastorff et al., 2025), countries’ level of digitalization may have influenced their ability to respond to the challenges of the crisis and therefore warrants further investigation.
Methods. To investigate changes in the influence of SES on reading and mathematics achievement we applied meta-analytic techniques to representative PISA data from 2018 and 2022, comprising 15-year-old students from 70 countries (N = 1,105,476; see Brunner et al., 2022). We tested whether survey year moderated the SES-achievement relationship and examined cross-country variability through random-effects models. Lastly, by conducting moderator analyses, we investigated whether the countries’ characteristics such as digital development (Digital Readiness Index; Cisco, 2024) and the length of school closures (Blavatnik School of Government, 2023) explain differences in the interaction effect, while simultaneously controlling for differences in economic strength (GDP per capita, World Bank, 2025).
Results. We found a positive association between SES and reading and mathematics achievement (ßReading = .32, 95%CI [.30, .34]; ßMathematics = .34, 95%CI [.31, .36]). The results indicated that there was overall variability in the examined effects between countries (QReading(69) = 434.88, p < .0001, I²Reading = 82.81%, τ²Reading = 0.006; QMathematics(69) = 547.81, p < .0001, I²Mathematics = 86.12%, τ²Mathematics = 0.009). Notably, the summary interaction effect between SES and survey year was not significant (ßReading = .015, 95%CI [-.004, .033]; ßMathematics = .016, 95%CI [-.005, .038]). The interaction model still showed high variability in the effects between countries (QReading(69) = 450.59, p < .0001, I²Reading = 73.82%, τ²Reading = 0.004; QMathematics(69) = 468.62, p < .0001, I²Mathematics = 80.49%, τ²Mathematics = 0.006). Notably, only the DRI significantly moderated the interaction for reading (ßReading = 0.03, 95%CI [.01, .06], p = .01) and mathematics (ßMathematics = 0.05, 95%CI [.02, .07]; p = .001), suggesting that SES-related disparities were more pronounced in further digitalized countries.
Discussion. This study shows a significant overall effect of SES on achievement, but no significant overall interaction effect of survey year and SES on achievement. We extend prior work (e.g. Betthäuser et al., 2023; Kastorff & Heine, 2024) by demonstrating that the national context, particularly the level of digitalization, shaped how SES-related achievement disparities evolved during the pandemic. Therefore, we plan to conduct further analyses focusing on differences in equity-focused policies to help us better understand how past disruptions can inform the development of more resilient and equitable systems.

Authors

• Maren Müller, Technical University of Munich

• Jacqueline Carbajal, Technical University of Munich

• Lena Keller, Kiel University (CAU)

• Tamara Kastorff, Technical University of Munich