Assessment instruments (the spine)
| Instrument | Cycle | Measures | Poland |
|---|---|---|---|
| PISA | 2022 | 15-year-olds — math / reading / science | 489 / 489 / 499 |
| TIMSS | 2023 | Grade 4 — math / science | 546 / 550 |
| PIRLS | 2021 | Grade 4 — reading | 549 |
| PIAAC | 2023 | Adults 16–65 — literacy / numeracy | 236 / 239 |
OECD averages used as anchors — PISA 2022: math 472, reading 476, science 485; PIAAC 2023: literacy 260, numeracy 263. TIMSS/PIRLS scales are centred at 500 by construction. PISA 2018 Poland baselines (confirmed via IBE): math 516, reading 512, science 511.
Context indicators
| Indicator | Poland | Anchor | Source |
|---|---|---|---|
| Public education spending, % GDP | 4.1% | 4.7% (OECD) | OECD Education at a Glance 2025 |
| Teacher statutory salary | rank ~38/44 | among lowest OECD | Education at a Glance 2025 |
| Early-childhood participation (3+) | 96.6% | 90.3% (EU) | EU Education & Training Monitor 2025 |
| Adult participation in learning | 20.3% | 39.5% (EU) | EU Education & Training Monitor 2025 |
| Equity — ESCS math gap | ~96 pts | ~93 (OECD) | PISA 2022 Vol I (snippet-sourced) |
School model, sector and homeschooling
Figures used in the “School model” section. Full caveats — including the list of numbers that failed verification and therefore never reached the site — are in docs/school-models.md in the repository.
| Indicator | Poland | Anchor | Source |
|---|---|---|---|
| Compulsory instruction time (grades 1–8, total) | 5 304 h | 7 642 h (OECD) | OECD Education at a Glance 2025 |
| Teacher autonomy — choice of methods | 95% | 94% (EU) | IBE, TALIS 2024 (grades 5–8, self-reported) |
| Autonomy — curriculum flexibility / assessment | 68% / 74% | 76% / 82% (EU) | IBE, TALIS 2024 (grades 5–8) |
| Influence on policy / feeling valued | 9% / 10% | 15% (EU) / 35% (OECD) | IBE, TALIS 2024 |
| Creative thinking (age 15) | 34 pts | 33 (OECD), 41 (Singapore) | OECD PISA 2022 (Creative Thinking) |
| Curriculum slimmed down | ~20% | 18 subjects, from 1 Sep 2024 | Polish Ministry of Education |
| Pupils in non-public schools | ~7,4% | 7.5% (Eurostat, same national returns) | Statistics Poland 2024/25 — own calculation |
| Pupils in homeschooling | ~1.2–1.3% | 96.7% in non-public schools | Polish school information system / Ministry |
| 1999 reform — delayed selection | ~100 PISA points | for vocational-track pupils | Jakubowski et al. (quasi-experimental) |
| Six-year-olds among first-graders | ~1.2% | 5,033 of 428,512 (upper bound) | Statistics Poland 2024/25, tables 3.1 + 3.4 — own calculation |
Tracks and foreign languages
Figures used in the “Real divide” section, read from the GUS tables rather than the press release. The track split is an enrolment stock, not a first-year intake — programmes differ in length, so the stock under-weights the short track. The Eurostat figure is not independent confirmation, since it restates the same GUS returns; its value is that every country is counted the same way. Full caveats in docs/enrichment.md.
| Indicator | Poland | Anchor | Source |
|---|---|---|---|
| Post-primary pupils by school type | liceum 44.7% / technikum 40.5% / vocational-entry and other 14.8% | n = 1,699,834, schools for youth | Statistics Poland 2024/25, table 4.1.1 |
| Vocational-track share | 54,5% | 54.4% (Eurostat, same national returns) against 49.9% EU-27 | Statistics Poland — own calculation; Eurostat educ_uoe_enrs04 |
| Foreign languages per pupil by school type | technikum 2.00 / liceum 1.99 / stage-I vocational 1.02 | primary 1.27 (2nd language from grade 7) | Statistics Poland 2024/25, table 5.1 — own calculation |
| Matura pass rate by school type | 2024: liceum 92.5 / technikum 85.2 / stage-II vocational 28.0 | 2025: 90.4 / 79.5 / 24.3 (only 1,353 sat stage-II) | CKE Sprawozdanie, table 5 and note 5, p. 14 (post-resit) |
Reproducible backbone
The machine-readable dataset lives at data/indicators.json in the repository, with provenance per value. PISA-math standing was independently recomputed from the Our World in Data grapher CSV (which republishes the OECD table): OECD n = 37, mean 472.4, SD 34.0, Poland 489.0, giving z = +0.49. The SD-unit recompute and the age-band cohort test were reproduced by an independent verifier against the raw tables.
Illustrations
The drawings behind the home-page sections were computer-generated with Google’s Gemini 3.1 Flash Image model (gemini-3.1-flash-image), in the style of a pen and pencil sketch on graph paper. They are decoration, not data: they encode no values and depict no real people or events. Only the charts carry numbers. Every file was inspected at full resolution before publication. The prompt, model, date and SHA-256 hash of each drawing are recorded in the repository at site/src/assets/sections/manifest.json.