Milos Kankaras | 2026 | Country note
The United Arab Emirates recorded its largest PISA gains in 2025. This note asks how much of that gain is improved learning rather than a changed test population, what separates the country’s very different school systems, and how far the measurement itself can be pushed.
The national average covers seven largely separate school systems. In 2025 British-curriculum private schools averaged 520.6 in mathematics and public Ministry schools 373.0, a distance of 147.6 points that has barely moved since 2022. Separating that gap into its measurable parts puts about 60 percent of it down to who the pupils are and who else is in their school, and under 5 points down to attendance, classroom climate and reported school resources together.
An item-level audit of the Arabic and English versions of the test finds that between three and four items in ten behave differently in Arabic, by a substantial margin each, but that the differences run both ways and cancel. Correcting every one of them moves the Arabic mean by at most one and a half points. The national mean, the sector means and the language gap can be reported as they stand; the subscale profiles a ministry would use to diagnose a curriculum weakness cannot.
The note states the limit of each analysis alongside its result. It attributes the improvement to no reform, reports no result for Emirati pupils because PISA records birthplace rather than citizenship, and does not claim that the language audit invalidates any published score.
Download the note as a PDF, 34 pages with 24 figures. Download the figures and data package for the exact plotted values, an Excel workbook with live charts, the methods annex and the list of published sources.
Part of the study The United Arab Emirates in PISA 2025.