Scoring · raw marks · equating

TMUA grade boundaries, explained.

There is no permanent table that turns a live raw mark into a TMUA score. Historical tables are real. Current conversions are form-specific and are not published as advance boundaries.

The direct answer

TMUA does not have fixed grade boundaries like an A-level paper. The same raw total is not guaranteed to produce the same reported score on every test form.

UAT-UK says it does not provide grade boundaries or a pass mark. Candidate responses are analysed using an item-response model, differences in form difficulty are removed through equating, and Papers 1 and 2 are combined into one overall score.
Official process Paper-specific history Practice estimate

From forty answers to one reported score

The raw scoring itself is simple. Each paper contains 20 multiple-choice questions. Correct answers earn credit, incorrect answers do not lose marks, and the two papers are used together. What happens after that is the part that makes a universal conversion table impossible.

01 · ResponsesCorrect and incorrect answers

Your responses across both 20-question papers are recorded.

02 · DifficultyItems are placed on a difficulty scale

The Rasch model estimates item difficulty and candidate ability.

03 · EquatingForm differences are removed

Small differences between versions are adjusted so candidates share one ability scale.

04 · ReportingOne score from 1.0 to 9.0

The common ability estimate is mapped to a readable overall TMUA score.

This does not mean the mark is arbitrary. It means a raw mark must be interpreted in the context of the questions that produced it. A slightly harder form should not disadvantage its candidates merely because it contains more difficult items.

Three different things are often called a boundary

A published conversion for one released paper

The released papers from the earlier archive include year- and paper-specific conversion information. It is valid for marking that particular historical paper. It is the best way to compare your attempt with the scale used for that paper.

Use it forMarking the named released paper
Do not use it forPredicting an unreleased live form exactly
ConfidenceOfficial for that historical paper

Why the same raw mark can mean something different

Imagine two forms that both contain 40 questions. On Form A, several items are answered correctly by most candidates. On Form B, those positions contain more demanding items. A raw score of 25 on each form does not automatically show identical ability. Equating exists to place both performances onto a comparable scale.

Illustration, not a conversion tableEqual raw totals can come from unequal forms

No invented scaled scores are attached to this example.

Form A25 / 40

Raw total observed on one set of items.

Form B25 / 40

The same total observed on a different set of items.

EquatingDifficulty is considered

The forms are placed on one ability scale before reporting.

Can you compare old and current TMUA scores?

You can compare them cautiously, but not by pretending the distributions are unchanged. The reported range remains 1.0 to 9.0, yet the current explanation fixes the typical point around 4.5 and places approximately 10% of candidates above 7.0. Older paper conversion tables were produced for older cohorts and administrations.

For preparation, keep the historical score attached to the paper and year. If you also calculate a current-style estimate, display the two numbers separately. That is why Euclid's converter describes the historical result as official for its paper and the current equivalent as an estimate.

Never write “14 marks equals 6.5” without naming the paper. A raw mark has no honest scaled meaning until the form and conversion method are identified.

How to mark a mock without fooling yourself

  1. Start with the raw mark. It is the least ambiguous measure of that attempt.
  2. Use the conversion belonging to the paper. Do not borrow a table from an easier or harder year.
  3. Separate correct guesses. A lucky answer should remain in the score but be flagged for review.
  4. Record timing. An untimed 26 and a timed 26 answer different preparation questions.
  5. Use a rolling median. Three to five recent papers provide a more stable picture than your best attempt.

If your aim is improvement, the error categories matter more than a decimal-place prediction. A score converter can organise the result. It cannot replace careful review.

Sources and accuracy policy

Use the right table

Convert the paper you actually sat.

Select a released paper, enter the raw marks and keep the official historical score separate from any current-scale estimate.

Use the TMUA score converter
Published and checked 1 August 2026Scoring method checked against current UAT-UK guidance