# Data sources ## School locations, sector, denomination, address - **Source**: "Schools_Scotland__2022" ArcGIS FeatureServer layer, published by the Scottish Government (derived from the "School contact details" and "School level summary statistics" publications on gov.scot). - **URL**: `https://services-eu1.arcgis.com/ELpYE44CpoxrJqcU/ArcGIS/rest/services/Schools_Scotland__2022/FeatureServer/0` - **Fetched**: 2026-07-07 - **License**: Open Government Licence v3.0 - **Raw files**: `data/raw/schools_scotland_page1.geojson`, `data/raw/schools_scotland_page2.geojson` (2 pages of up to 2000 records each, `outSR=4326` for WGS84 lat/lng). - **Notes**: covers all ~2,495 schools in Scotland; filtered at build time to primary/secondary sectors only (special schools excluded), giving 2,360 schools across all 32 local authorities. ## Primary school ratings (inspection quality indicator grades) - **Source**: Scottish Government FOI response FOI-202000044014, Annex A spreadsheet. - **URL**: `https://www.gov.scot/publications/foi-202000044014/` (spreadsheet linked from that page) - **Fetched**: 2026-07-07 - **Raw file**: `data/raw/foi_inspection_annex_a.xlsx` - **Coverage**: inspections from April 2008 to June 2020, across three inspection framework periods each grading a different set of quality indicators (QIs) on a 1 (weakest) to 6 (strongest) scale: - 2008-2016: QI 1.1, 2.1, 5.3, 5.1, 5.9 - 2016-2020: QI 1.3, 2.3, 3.2, 3.1 - 2017-2020: QI 1.1, 2.3, 3.2 - **Known limitation**: no confirmed bulk source was found for inspections after June 2020. Scotland's inspection model is light-touch (only a fraction of schools inspected in any given period), so schools with no rating in the map may simply not have been inspected recently, and schools that do have a rating may be reflecting an inspection from as far back as 2008. The `inspectionDate` shown in each popup is the actual date of the most recent inspection Annex A contains for that school (deduplicated by SEED code, keeping the latest date where a school appears in more than one framework period) — always check it rather than assuming the rating is current. - **What "rating" means here**: Education Scotland does not publish a single official overall grade per school (unlike Ofsted's England-only "Outstanding/Good/..." scale). The `label` and `averageScore` fields in `data/schools.json` are an illustrative average computed across whichever quality indicators were graded at that school's inspection — not an official Education Scotland score. - **Usage**: this is the sole rating source for primary schools, and the fallback for any secondary school with no attainment data available (see below). ## Secondary school ratings (SQA attainment) - **Source**: Scottish Government, "Schools - Breadth and Depth of Qualifications", statistics.gov.scot (a linked-data cube, OGL v3.0 licensed). - **URL**: `https://statistics.gov.scot/data/breadth-and-depth`, queried via its SPARQL endpoint at `http://statistics.gov.scot/sparql`. - **Fetched**: 2026-07-09, filtered to: `numberOfAwards = 5-and-above`, `scqfLevel = 6-and-above` (SCQF level 6 = Higher), `courses = all-courses`, `comparator = real-establishment`, for reference periods 2023-08-01/P1Y (academic year 2023-24) and 2024-08-01/P1Y (2024-25). - **Raw files**: `data/raw/breadth_and_depth_2023-24.json`, `data/raw/breadth_and_depth_2024-25.json` (raw SPARQL JSON results). - **Metric**: % of the school-leaver cohort attaining 5 or more qualifications at Higher level (SCQF 6) or above. This is a genuinely school-level, annually-updated official statistic — unlike primary attainment (ACEL), which Scottish Government only publishes aggregated at local-authority level, so no equivalent exists for primary schools here. - **Join key**: each establishment entity's `foi:code` in this dataset is the same SEED code used throughout the rest of this project — confirmed by cross-checking real schools (e.g. SEED 5102030 = "Thurso High School" in both this dataset and the ArcGIS schools layer). - **Suppression**: cells marked `*` (small-cohort disclosure control) or `#` (not applicable) in the raw data are treated as no data for that year; 2024-25 is preferred where both years have a usable value, otherwise 2023-24 is used. - **Banding**: `data/schools.json` reuses the same six labels as the primary inspection scale (Excellent/Very Good/Good/Satisfactory/Weak/Unsatisfactory) so the existing "School rating" filter needs no changes, banding the raw percentage at 60/45/35/25/15/0. These thresholds are illustrative (chosen from the actual national distribution, not an official scale) — Scottish Government does not publish rating bands for this metric. - **Known limitation**: this is a raw attainment figure, not adjusted for pupils' socioeconomic background or prior attainment. The official methodology recommends comparing a school against its own "virtual comparator" (a matched sample with similar pupil characteristics) for a fairer read — this map does not show that comparison, only the raw percentage, and the popup discloses this. ~16 of 360 secondary schools have no usable value in either year (suppressed in both, or absent from the dataset) and fall back to inspection-based ratings where available. ## Region-picker boundary shapes (index.html) - **Source**: ONS Open Geography Portal, "Local Authority Districts (December 2023) Boundaries UK BUC" (Ultra Generalised, 500m — the lightest-weight ONS boundary product, intended for small-scale thematic maps like this one). - **URL**: `https://services1.arcgis.com/ESMARspQHYMw9BZ9/arcgis/rest/services/Local_Authority_Districts_December_2023_Boundaries_UK_BUC/FeatureServer/0` - **Fetched**: 2026-07-09, filtered to Scotland (`LAD23CD LIKE 'S12%'`), projected in British National Grid (EPSG:27700). - **Raw file**: `data/raw/scotland_la_boundaries.geojson` (all 32 Scottish local authorities). - **License**: Open Government Licence v3.0 / © Crown copyright and database right, ONS. - **Notes**: ONS spells the Outer Hebrides council "Na h-Eileanan Siar"; the schools dataset (and `js/regions-data.js`) uses "Na h-Eileanan an Iar" for the same area — both refer to the same local authority, reconciled in `scripts/build_regions_geometry.py`. ## Refreshing the data Re-run `python3 scripts/build_schools_json.py` after replacing the files in `data/raw/` with newer downloads from the same sources (or newer equivalents) to regenerate `data/schools.json`. Re-run `python3 scripts/build_regions_geometry.py` after replacing `data/raw/scotland_la_boundaries.geojson` to regenerate `js/scotland-geometry.js`. To pull a newer year of secondary attainment data, query the SPARQL endpoint with the reference period updated to the new year (e.g. `2025-08-01T00:00:00/P1Y` for 2025-26), and save the result as `data/raw/breadth_and_depth_.json`: ``` curl -sS -G "http://statistics.gov.scot/sparql" \ --data-urlencode 'query=PREFIX qb: PREFIX sdmxd: PREFIX foi: PREFIX edu: PREFIX dim: PREFIX measure: SELECT ?seedCode ?schoolName ?refPeriod ?percent WHERE { ?obs qb:dataSet ; sdmxd:refPeriod ?refPeriod ; edu:refEstablishment ?est ; dim:numberOfAwards ; dim:scqfLevel ; dim:courses ; dim:comparator ; measure:percent ?percent . ?est foi:memberOf ; foi:code ?seedCode ; foi:displayName ?schoolName . FILTER(?refPeriod = ) }' \ -H "Accept: application/sparql-results+json" -o data/raw/breadth_and_depth_YYYY-YY.json ``` Then update the `sources` list at the top of `load_attainment()` in `scripts/build_schools_json.py` to include the new filename (most recent year first).