How the figures are checked
A result is easier to understand when you can see who published it, what it counts and whether it was later revised. We keep those details with the figure. When the evidence supports a correction, we explain the change and preserve the earlier value. When publications still disagree, you can see both and how we treat the difference. Read the corrections and published differences
The letters used in our source notes
| Code | Meaning |
|---|---|
| CONFLICT | Publications disagree, and we have not yet resolved the difference. |
| D | A school’s destinations booklet or list. |
| P | An original school, examinations or Freedom of Information publication. |
| R | Calculated from other information, using stated assumptions. |
| S | A press report, league table or other publication quoting results second-hand. |
Definitions
What the terms mean
- Acceptances
- Offers a pupil has firmly or finally accepted, using the definition in the UCAS or university table. The admissions round is always named.
- Applications
- The pupils or application records submitted, as the source counts them. An application tells us that someone applied; it does not tell us the outcome.
- Which years a head’s tenure covers
- A pupil leaving in year Y may have joined around Y−5, taken GCSEs in Y−2 and applied through UCAS in Y−1. Before linking results to a head’s tenure, we need to know who led the school throughout the relevant years.
- Destinations
- The places named in a school’s leavers’ list or booklet. A list becomes a rate only when we also know how many leavers that same list covers.
- Matriculations
- Pupils who enrolled at a university. The total can differ from acceptances because of deferred entry, unmet conditions or a change of plan.
- Missing figures
- A missing figure stays blank. We keep a year even if only one school has a result; we never invent the other school’s figure.
- Offers
- Conditional or unconditional offers from a university. They tell us about an admissions decision, before we know whether a pupil will take up the place.
- The pandemic years
- In 2020 and 2021, grades were assessed by schools and teachers after exams were cancelled. These years are shown separately from ordinary examined-year trends.
- One admissions round
- Applications, offers and final acceptances from one UCAS or university admissions round. We keep these together, apart from leavers’ destinations or offers from a round still in progress.
Grade mappings
Grade mapping
GCSE
Old A*–G: A*, A*–A, A*–B. New 9–1: grade 9 has no exact old-scale equivalent; 9–8 is the nearest old-A* top-equivalent; 9–7 is A*/A-equivalent; A*–B requires grade 6.
Grade mapping
KCS Wimbledon’s combined A-level and IB headline
For the internal combined KCS headline only: IB 7→A*, 7–6→A*/A, 7–5→A*/B.
Grade mapping
PRE U
On TSR’s comparison scale, Pre-U D1–D2 counts as A* or equivalent, D1–D3 as A*–A or equivalent, and D1–M2 as A*–B or equivalent.
The checks behind a comparison
- Every chart says what it counts, which years it covers and where the figures came from. It identifies the total behind a percentage and any limits in the evidence.
- A comparison starts with figures measuring the same thing. If a chart adds a different measure for context, its label and appearance make the distinction clear.
- When publications still disagree, you can see both figures and why we have not chosen between them.
- Once a figure has been corrected, comparisons use the correction. The earlier headline stays in the history.
- The full tables in v13 remain available where the shorter v14.6 account omits rows.
- The school-assessed grades of 2020 and teacher-assessed grades of 2021 stay separate from ordinary examined years.
- Where Cambridge’s figures lack a breakdown, we leave it missing. We also leave the SPGS 2025 A*/A figure blank unless a source supplies it.
- Winchester’s comparison series uses the approved Pre-U publications for 2011–2019, with each publication’s scope and precision retained. Earlier versions remain in the source record. The teacher-assessed results for 2021 remain outside the default comparison.
Which research version we use
Our research has grown through several versions. A newer summary may leave out detail that an earlier table contains, so the date alone does not decide which account we use. The table below explains what each version contributes.
| Order | Version | What we use it for | Where another source is needed |
|---|---|---|---|
| 1 | v14.6 | The seven-school summary, corrections after v14 and archived KCS and Westminster destination lists | Detailed annual tables omitted from the summary still come from earlier versions. A later second-hand figure does not replace a count checked against an original source in v13.9. |
| 2 | v13 consolidated reader | Complete admissions-round series and detailed annual tables omitted from v14.6 | — |
| 3 | v13.9 master + addendum | Winchester figures checked against original table images and grade counts, with the calculations based on them | — |
| 4 | v14.0–v14.2 addenda | Differences checked against the prospectus, GCSE corrections and evidence from league tables. | — |
| 5 | legacy attachment | Earlier research kept so changes can be followed. It does not override a later correction checked against a source. | — |
How the research developed
Early collected results
The first collection of Winchester, Eton, St Paul’s and KCS figures. Later research has replaced individual figures where better evidence was found.
Bringing the research together
Earlier accounts brought together, with fuller KCS material and additional checks between sources.
Finding earlier sources
New ways into archives and leads for historical research. Most of these additions did not change the downloadable figures.
Following results over time
Long-term rates, an examination of what contributed to falling results, and comparisons using two separate measures.
Leadership and the years pupils spent at school
A record of headmasters’ tenures and the time between pupils joining a school and receiving their results.
Checking original tables
Winchester’s 2018/2019 table images, the correction to its 2019 GCSE results, separate treatment of pandemic years and older league-table comparisons.
Further corrections
Offers and enrolments checked against the prospectus; GCSE results for 2016/2019 checked against original sources; and a separate account of the D1–M1 figure in a third-party league table.
Bringing seven schools together
Westminster, KCS, SPGS and Wycombe added to the account, GCSE corrections completed and archived destination lists checked. Version v14.6 is the latest summary, but earlier detailed tables are still needed.
Oxford and Cambridge records
| Acceptances and admissions | Oxford uses labels including “Accepts” and “students admitted”. Cambridge counts unconditional firm acceptances after confirmation. These are later outcomes than offers, so their rates are labelled separately. |
|---|---|
| Applications | Completed applications from the group the source names. |
| How we identify the admissions year | Oxford’s FOI/202505/554 release says UCAS Cycle 2025 covers entry in 2025 or deferred entry in 2026; it uses the same convention for 2022–2024. Cambridge also names rounds for the intended start year. KCS’s archived January 2013 announcement exactly matches the Oxford and Cambridge offers for round 2013. We therefore use the year named by the university, without adding a year. |
| Admissions year | The university names the round for the intended year of entry. Cambridge explains that applications begin in September of the previous calendar year. |
| Destination | The university a school leaver eventually attends. It is a different outcome from an application, offer or acceptance. |
| Intended year of entry | The year a pupil intends to begin university. A deferred-entry applicant may start one year later. |
| Offers | Conditional or unconditional offers, unless the source gives a narrower definition. |
| When small counts are withheld | Small counts are sometimes withheld. In these Oxford tables, <3 or a blank indicates an unknown count below 3. Older Cambridge releases use <4 or <3. One unidentified Cambridge 2025 source reportedly rounds counts of 1–4 up to 5. We retain these limits and uncertainties; we never replace them with a midpoint. |
US and overseas university records
- Totals are labelled. Adding a total to its individual university rows would count the same results twice.
- An offer, an acceptance, a firm place and a final destination tell us different things. A university named without a count tells us less again. We keep these distinctions visible.
- If a university is named without a count, we know only that it appears in the list. We do not assume that 1 pupil went there.
- A total for the US covers a different area from North America or all overseas destinations. Each keeps its own heading.
- Where versions of a source disagree, the figures stay out of comparisons until the difference is resolved.