This study develops and applies a statistical framework for analyzing Danish PISA data, emphasizing the role of demographic and socioeconomic determinants in student achievement. Drawing on PISA 2022 (mathematics) and PISA 2018 (reading). The core multilevel model specifies achievement as a function of gender, immigrant background, individual socioeconomic status (ESCS), and school-level ESCS. Results reveal systematic disparities: boys outperform girls in mathematics, whereas girls exhibit superior reading performance; students of Danish origin consistently achieve higher scores than peers with immigrant backgrounds. ESCS exerts a strong positive influence, particularly among Danish-origin students, while contextual effects of school-level ESCS remain significant.
Robustness checks—including extended interaction models, diagnostic analyses of residuals, and correlations between peer effects and school-level teaching effects—confirm the stability of the core model. Additional analyses employing single imputation indicate only marginal changes in estimated parameters.
Extensions incorporating PISA-derived indices markedly improve model fit: mathematics anxiety (ANXMAT) nearly doubles explained variance for mathematics, while reading enjoyment (JOYREAD) significantly enhances predictions for reading. Findings underscore persistent structural inequalities and highlight the necessity of integrating psychosocial constructs into large-scale assessment research. This framework contributes to debates on equity, inclusion, and interpretation of international assessment data.

Authors:
Hans Bay, Copenhagen University, Denmark

About the Presenter(s)
Hans Bay is Associate Professor at the University of Copenhagen with over 40 years of experience in applied statistics and survey methodology. He has led major international assessments, including PISA and ESS, and his research focuses on statistical

Connect on Linkedin
https://www.linkedin.com/in/hans-bay-7b4b015

See this presentation on the full schedule – Saturday Schedule

Bookmark