How reading is marked

Reading marking is a pure function. No model is involved, the answer key is fixed, and the same answers would produce the same score every time.

How your answer is compared

Every question's key holds every accepted form of the answer, so both "cotton" and "raw cotton" can be right where both are.

Comparison ignores capitalisation and surrounding whitespace. Typing TRUE is not a mistake and is not marked as one.

Raw score to band

Your raw score is scaled up to 40 first, so a 13-question practice set reports on the same scale as a full test. Then:

Raw score, scaled to 40Band
39 or more9.0
37 or more8.5
35 or more8.0
33 or more7.5
30 or more7.0
27 or more6.5
23 or more6.0
19 or more5.5
15 or more5.0
13 or more4.5
10 or more4.0
8 or more3.5
6 or more3.0
Below 62.5

Academic scale, both formats

This is the public Academic Reading conversion, and it is currently applied to General Training candidates too. General Training converts more generously, so if you are sitting General Training, treat your reading band as conservative until that ships.

The review

The review page shows:

  • Your band, and the raw score behind it, labelled as an estimate rather than a score.
  • Every question, marked correct or incorrect.
  • For each one, the evidence sentence — the line in the passage that makes the answer the answer — and an explanation of why.
  • Recurring patterns: question types you are consistently weak at across every attempt you have made, not just this one.

The evidence sentence is the part worth slowing down on. Getting a question wrong tells you little; seeing the sentence you needed to have found, and realising you read past it, tells you what to do differently next time.