If your child has sat an educational assessment recently — an 11+ mock, a CAT4 test at school, or an online diagnostic — you have probably encountered a “standardised score” alongside the raw percentage. Most parents ignore the standardised score and focus on the percentage. That is the wrong instinct. Here is what a standardised score actually means, why it tells you far more than a percentage, and how to interpret it for decisions about your child's education.
The Problem with Raw Scores
A raw score — say, 43 out of 60 — tells you one thing: your child answered 72% of the questions correctly on this particular test on this particular day. It does not tell you whether the test was easy or hard. It does not tell you how other children at the same age performed. It does not tell you whether 72% represents strong performance or weak performance for a child of that age. And it does not allow any meaningful comparison between this test and a different test your child takes in three months.
These are precisely the questions that parents preparing for selective admissions — or simply trying to understand how their child is progressing — need to answer. Raw scores cannot answer them. Standardised scores can.
A standardised score transforms a raw score into a number that accounts for age, test difficulty, and the performance of a reference population. Two children who both answer 43 out of 60 questions correctly may receive very different standardised scores if one is 18 months older than the other, or if they sat tests of different difficulties. The standardised score puts them on a level playing field — comparing each child only against others of the same age.
The Standardised Scale
Almost all UK standardised assessments — including GL Assessment's 11+, CAT4, NFER, and Eduentry — use the same scale: a mean (average) of 100 and a standard deviation of 15. This is sometimes called the "standard score" scale and it is the same scale used by Wechsler IQ tests, the Stanford-Binet, and most professional psychometric assessments. This shared scale makes comparisons meaningful: a score of 115 on Eduentry and a score of 115 on a CAT4 test represent the same relative position within the same-age population.
| Score range | Classification | Approx. percentile | % of population |
|---|---|---|---|
| 130+ | Very Superior | Top 2% | ~2% |
| 120–129 | Superior | 91st–98th | ~7% |
| 110–119 | Above Average | 75th–91st | ~16% |
| 95–109 | Average | 37th–63rd | ~25% |
| 85–94 | Below Average | 16th–36th | ~16% |
| 70–84 | Low Average / Needs Support | 2nd–15th | ~14% |
An important property of this scale: because the standard deviation is 15, each 15-point step represents exactly one standard deviation. A score of 115 (one SD above the mean) is approximately the 84th percentile. A score of 130 (two SDs above the mean) is approximately the 98th percentile. These relationships are consistent across all assessments that use this scale.
What Is a Percentile?
The percentile is the most intuitive way to interpret a standardised score. Your child's percentile tells you what percentage of children of the same age they performed better than. A score in the 84th percentile means your child performed better than 84% of children their age — and was outperformed by 16%.
There are two common misconceptions about percentiles worth addressing. First: the 50th percentile is not a "bad" score — it means exactly average, better than half and worse than half. Many parents see a score in the 50th percentile and assume their child is struggling. They are not; they are performing at the population median. Second: percentile is not the same as percentage. A child who answers 70% of questions correctly may be in the 85th percentile if the test was hard, or the 30th percentile if the test was easy.
- Standardised score 13098th percentileTop 2% of same-age children
- Standardised score 12091st percentileTop 9%
- Standardised score 11584th percentileCompetitive grammar school range
- Standardised score 11075th percentileAbove average
- Standardised score 10050th percentileExactly average
- Standardised score 9025th percentileBelow average
What Is a Standardised Age Score (SAS)?
The SAS is the specific standardised score format used by GL Assessment in the 11+ exam. It introduces one additional adjustment: the child's exact age in months at the time of the test.
This matters because children taking the 11+ in September of Year 6 range in age from approximately 10 years and 2 months to 11 years and 1 month — a developmental difference that is significant at this age. Research consistently shows that the oldest children in a cohort outperform the youngest on standardised tests, not because of greater ability, but because of developmental advantage.
The SAS formula compares each child only against others born in the same month range — typically within a two-month window. A child born in August who answers 43 questions correctly may receive a higher SAS than a September-born child who also answered 43 correctly, because the bar is set relative to August-born children. The SAS effectively gives summer-born children a fair chance despite their younger age.
Practical implication for parents:If your child is summer-born, do not panic when a September-born classmate seems to be performing better in practice papers. The SAS adjusts for this. What matters is your child's performance relative to children born in the same month range — not relative to the whole cohort.
How Standardised Scores Are Calculated
For parents curious about the mechanics: most modern standardised assessments use Item Response Theory (IRT) to calculate scores, rather than simply counting the number of correct answers. IRT estimates the child's underlying ability level (often denoted θ, "theta") based on which specific questions were answered correctly and incorrectly — not just how many.
In a 2-Parameter Logistic (2PL) IRT model, each question has a difficulty parameter and a discrimination parameter. Getting a hard question right provides stronger evidence of high ability than getting an easy question right. Getting an easy question wrong is stronger evidence of low ability than getting a hard question wrong. After estimating theta from the pattern of responses, the score is converted to the standardised scale (mean 100, SD 15) using a linear transformation.
The practical implication: on an adaptive test like Eduentry, getting early questions right matters slightly more than it would on a fixed test, because early responses have greater influence on the initial theta estimate. This is why maintaining composure and accuracy in the opening questions — rather than rushing to "save time" for later — is important strategy.
What Score Is Needed for Grammar School?
Grammar school cut-offs vary by area, school, and year — but the following ranges apply as general benchmarks for 2026 entry:
- SAS 111–114: Borderline at many schools. May be placed on a selective register but below the competitive cutoff. Distance from school becomes the deciding factor.
- SAS 115–120: Comfortably within the selective range for most grammar schools in Kent, Essex, and Hertfordshire. Strong enough to be offered a place at most schools in these areas if within catchment.
- SAS 121–128: Required for the most competitive grammar schools, including those in Sutton, Buckinghamshire, and selective schools in Birmingham.
- SAS 128+: Needed for the most oversubscribed London schools — Queen Elizabeth's Boys (Barnet) and The Henrietta Barnett School — where demand far exceeds supply even at very high scores.
For a full breakdown by area, see our Grammar School Entry Requirements 2026 guide, which covers every major grammar school area in England with specific score benchmarks.
What to Do with Your Child's Score
When you receive a standardised score, the most important thing to do is resist interpreting it as a fixed characteristic of your child. A score at the 65th percentile today does not mean your child will be at the 65th percentile in 12 months. Standardised scores at this age are genuinely responsive to targeted preparation — particularly in verbal reasoning and maths, which are the most coachable subjects.
Use the score diagnostically. A child scoring below 95 in verbal reasoning and above 115 in mathematics needs a completely different preparation plan than a child scoring 108 across all four subjects. The percentile tells you where they are. The subject breakdown tells you what to work on. For a subject-by-subject breakdown of verbal reasoning specifically, see our verbal reasoning question types guide.
Re-test every 3–4 months to measure genuine progress. Month-to-month variation is largely noise. A 3–4 month gap allows enough time for real changes in performance to show in the score — and enough time for preparation to have had a measurable effect.