Overview
- PISA 2025 confirmed a large global drop in 15‑year‑old performance in reading, mathematics and science as reported in early September, a decline that stands regardless of how subgroups are measured.
- A reconstruction by Innovamat Research Lab with EsadeEcPol shows that PISA replaced book‑related items with material‑condition items in the possessions component of the socioeconomic index, which moved many students into different SES quartiles.
- Using a reconstructed index designed to match previous cycles, Innovamat finds the fall in scores is distributed similarly across SES groups rather than concentrated among the high‑SES quartile reported by the OECD.
- The OECD says its 2025 measurement choices were a valid methodological decision meant to reflect current markers of advantage, acknowledges comparability caveats in technical notes and invites further research, while independent statisticians dispute parts of Innovamat’s critique.
- The dispute has practical effects in Spain where Catalonia’s exclusion from the OECD report triggered internal investigations and parliamentary scrutiny, and several researchers have said they will amend papers that used the original PISA SES classifications.