“The school says my child is sitting a CAT4. Should we be practising?” That is usually the first question. The second comes a few weeks later, when a report full of SAS scores, stanines and “spatial bias” lands in your inbox and nobody has explained what any of it means.
Let me take both in turn. CAT4 is not like a maths test or an English paper. It does not check what your child has been taught. It looks at how they reason: with words, with numbers, with shapes, and with shapes turned around in their head. Schools use it to set classes and targets, and some independent schools use it for entry, especially at 16+.
There are four parts, which GL calls batteries. From Level A upwards, each battery has two short tests, so a child sits eight tests in all. The times below are for the digital edition and do not include the instructions and practice questions before each one.
Verbal Classification: three words that belong together, and you pick a fourth. Rose, tulip, daisy go with lily, not leaf.
Verbal Analogies: two words linked in some way, then a second pair to finish.
Number Analogies: pairs of numbers that follow one rule. In [3 → 12] [5 → 20] [8 → ?] the rule is “times 4”, so the answer is 32.
Number Series: find the next number in a sequence. Rough paper is allowed for both.
Figure Classification: three shapes that share something, and you find the one that matches.
Figure Matrices: a shape changes in a certain way, and you apply the same change to a second shape.
Figure Analysis: a square of paper is folded, a hole is punched, and you picture it unfolded.
Figure Recognition: spot a shape hidden, at exactly the same size, inside a busy drawing.
“This isn’t about what you know. It’s about spotting the rule.”
Every CAT4 question has one rule hiding in it. Find the rule, check it works every time, then pick the answer.
CAT4 has 12 levels, each written for a different age. In England, GL recommends:
Schools can move a level up or down. GL gives two examples: a Year 6 group tested at the end of the summer term may sit Level D, and a selective school with a very able group may use the level above. If a school tells you the level, believe the school.
Each battery gets a Standard Age Score (SAS). The average is 100, and your child is compared only with children of the same age. About two thirds of children score between 85 and 115. Scores run from about 60 to 140.
The report also gives a National Percentile Rank. A rank of 84 means your child scored the same as or better than 84% of children their age. In practice an SAS of 115 is roughly that: comfortably above average, with about one child in six scoring higher.
Then there are stanines, a simple 1 to 9 scale:
So what counts as “good”? Honestly, it depends on the school. Ask each school what it looks for, and remember it reads your child’s interview, reports and grades alongside the numbers.
CAT4 also compares the verbal score with the spatial score and gives one of seven profiles, from “extreme verbal bias” through “no bias” to “extreme spatial bias”. Most children have no bias. It is a hint about how a child prefers to learn, not a label.
For older children, the report can also show GCSE and A-level indicators. These are chances, not promises: the most likely grade, a grade “if challenged”, and the probability of reaching, say, grade 7 or above.
Read a CAT4 report as a starting point, not a verdict.
It shows one test, on one occasion. GL itself says results are a snapshot and should never be seen as fixed ability. Effort, good teaching and confidence still decide far more than any single score.
GL says children do not need to prepare, and I agree that there is nothing to revise. There is no syllabus and no list of facts. What I do want is for a child to have seen each of the eight question types once before the day, so the first few minutes are spent thinking, not working out what the screen wants.
These are the two mistakes I see most often:
CAT4 questions are different every time, so there is nothing to memorise. A tired, bored child does worse, not better.
Every test is strictly timed and simply ends. A child who stalls on one hard question can lose the five easy ones after it.
This is how I would get a child ready:
A few examples of all eight, talked through together.
For analogies: “a glove goes on a hand”, so a sock goes on a foot. Then check the same sentence works for the answer exactly.
2, 5, 11, 23, ? The gaps are 3, 6, 12, so each gap doubles. The next gap is 24, and the answer is 47.
Check shape, then size, shading, number of sides, position, rotation and reflection, crossing out options as you go. Often two things change at once.
Fold a square, punch a hole, guess the pattern, then open it and check. It makes the spatial questions far less strange.
A wrong answer loses nothing. If a question is taking too long, make your best choice and move on.
Short and often wins. Fifteen minutes, a few questions, every mistake talked through.
Then one timed practice near the end, so the clock holds no surprises.
Yes, at some boarding schools. Bryanston says its sixth form candidates sit GL Assessment’s CAT4, and Malvern sets an online CAT4 alongside an English test. My 16+ page explains how I help students preparing for sixth form entry.
If your child is sitting other tests too, my guides to the ISEB Pre-Test and verbal reasoning question types cover a lot of the same skills, and the school finder shows how entry works at each school.
For children aged about 8 and over it is three parts of up to 45 minutes each, about 2 hours 15 minutes in all, including instructions and practice questions. The youngest children (about 6 to 8) sit two shorter parts of about 30 minutes.
No. CAT4 compares your child with children of the same age across the UK. Each school decides for itself what scores it looks for, and CAT4 is usually only one part of its decision.
Both exist. GL Assessment publishes a paper edition and a digital one for computers and tablets. Some schools, such as Malvern at 16+, set it online.
A simpler version of the score, from 1 (very low) to 9 (very high). Stanines 4, 5 and 6 are average and cover an SAS of 89 to 111. Stanine 9 means an SAS of 127 or more, the top 4% of children of that age.
No. The raw score is simply the number of questions answered correctly, which is then converted into a score for your child's age.
It can affect the verbal part. GL Assessment points out that the non-verbal and spatial parts do not rely on reading or English, which is why they are useful for children with English as an additional language.
I prepare children 1:1 online for 11+, 13+ and 16+ entry to the UK’s leading independent and grammar schools. Tell me about your child and the schools you are considering, and I will tell you honestly what the preparation should look like.
Enquire About LessonsSources: GL Assessment: CAT4 · GL Assessment: CAT4 teacher guidance, UK edition (PDF) · GL Assessment support: CAT4 quick data guide · GL Assessment support: understanding your CAT4 data · Bryanston: Sixth Form entry · Malvern College: Sixth Form admissions. Details change from year to year, so always check the official websites before you apply.