What CAT4 is, and is not
CAT4, in full the Cognitive Abilities Test fourth edition, is published by GL Assessment and measures reasoning ability across four domains: verbal, non-verbal, quantitative and spatial. It is designed as an aptitude test, intended to indicate how well a child can think, rather than what the child has already been taught. The distinction matters. A child who has missed a year of formal maths because of a move can still produce a strong quantitative reasoning score, because the test measures the underlying thinking skill rather than recall of mathematical content. Equally, a child with strong subject knowledge but weaker reasoning skills can score lower on CAT4 than their school grades would predict.
The test takes around two and a half hours and is delivered in three sittings, usually across two days. It is administered by the school online, with each child working at their own pace within timed sub-tests. Schools use it for two main purposes. First, to baseline a new arrival against a global cohort, particularly in the context of admissions or year-group placement. Second, to identify children whose attainment in subjects is materially below or above what their reasoning ability would predict, allowing the school to target intervention or extension as needed.
Standard Age Scores: the headline number
The most important figure on a CAT4 report is the Standard Age Score, often written as SAS. A SAS of 100 represents the population average for the child's age. The scale is built so that two-thirds of children sit between 89 and 111, and 95 percent of children sit between 78 and 122. The top 2 percent score above 127. The bottom 2 percent score below 73. These figures are stable across the test population in England, the country where CAT4 is normed, but international schools see broadly similar distributions when they test their full cohort.
The standard CAT4 report shows four SAS figures, one for each of the four domains, plus a combined Mean SAS that averages the four. The Mean SAS is the headline number schools use for at-a-glance reference. A Mean SAS in the high 110s and above is consistent with the strongest selective schools' incoming cohorts. A Mean SAS at 100 sits at the population average. A Mean SAS in the high 80s and below is consistent with a child likely to need additional support in mainstream curriculum work. The number is not a destiny. It is one data point among many.
Stanines and what they tell parents
The second figure parents commonly see on a CAT4 report is the stanine, short for "standard nine". This is a one-to-nine scale where 5 represents the population average, 9 is the top 4 percent and 1 is the bottom 4 percent. Stanines exist because they are easier to communicate verbally than three-digit SAS figures, and schools use them when talking with parents. A child with a verbal stanine of 7 and a quantitative stanine of 4 is, in plain English, materially stronger at language reasoning than at mathematical reasoning. A child with a stanine of 8 across all four domains is broadly able across reasoning types.
The structural advantage of stanines over raw scores is that they smooth out the noise of a single test sitting. A SAS of 109 and a SAS of 113 sit within the test's standard error of measurement, and treating them as meaningfully different is overinterpretation. The same two scores sit comfortably in stanine 7 and stanine 7, the same band. For working with the data, stanines are the right granularity. Save the three-digit SAS for the year-on-year comparison.
Want to understand the wider admissions picture?
CAT4 is only one of several tools schools use during admissions. Our admissions deadlines guide covers when in the calendar testing typically sits, and our document checklist walks through the wider paperwork. To shortlist schools that match your child's profile, use the school finder.
The four sub-tests: what each one measures
The verbal reasoning sub-test measures the ability to reason with words. Tasks include verbal classification (identifying which word does not belong), verbal analogies (completing patterns of words), and recognising relationships between word pairs. Strong verbal reasoning is correlated with strong outcomes in English, humanities and language subjects. A child with weak verbal reasoning but strong other domains often does well in subjects that are less language-dependent, and may benefit from explicit reading and vocabulary work.
The non-verbal reasoning sub-test measures pattern recognition with shapes. A child with strong non-verbal reasoning often performs well in mathematics, science and any subject that requires abstract pattern thinking. The non-verbal sub-test is the one least affected by English-language ability, which makes it particularly informative for children who have only recently moved to an English-medium school.
The quantitative reasoning sub-test measures the ability to reason with numbers, not arithmetic ability. Tasks involve number series, number analogies and equation building. Strong quantitative reasoning correlates with strong outcomes in mathematics and physics. A child with weak quantitative reasoning relative to their other domains may need explicit teaching of mathematical thinking, beyond standard curriculum mathematics.
The spatial reasoning sub-test measures the ability to manipulate shapes and form patterns mentally. Strong spatial reasoning is correlated with outcomes in design, engineering and certain branches of science and mathematics, particularly geometry. It is often the domain that surprises parents most, because it is rarely directly taught and so a strong score reveals capability that may not have shown up in school grades.
The profile picture: how the four scores fit together
The most useful read of a CAT4 report is not any single score but the shape of the profile across the four domains. Educators talk about three patterns. A "spiky" profile has noticeably different scores across the four domains, suggesting the child has clear cognitive strengths and weaker areas, and is often a strong candidate for differentiated teaching. An "even" profile has similar scores across all four, indicating broadly consistent reasoning ability across types of thinking. A "lopsided" profile, particularly one with strong non-verbal and spatial but weaker verbal and quantitative, can indicate a child for whom English is a second language, or where there is a specific learning profile such as dyslexia worth investigating further.
Schools use the profile picture in three ways. First, to identify children whose attainment is below what their reasoning ability would predict, which usually triggers a conversation about engagement, motivation or barriers to learning. Second, to identify children whose attainment is well above their reasoning profile, which typically indicates very disciplined effort and is recognised as such by good teachers. Third, to inform setting decisions, where schools group children by ability. The profile is the strongest predictor schools have at the start of an academic year, and a thoughtful school uses it to plan teaching rather than to label children.
What to do with a CAT4 report
If your child has just been CAT4-tested, you can do three useful things with the report. First, ask the school's pastoral lead or head of year to talk you through the four-domain profile and what they intend to do with it. Most schools will offer this conversation, particularly if the report shows a noticeable profile shape. Second, hold the data lightly. A single CAT4 sitting carries genuine measurement error, and re-testing two years later can shift individual sub-domain scores by half a stanine in either direction without anything substantive having changed. Third, treat the report as a conversation opener, not a verdict. A weaker domain is information to act on, not a fixed limit.
If your child has not yet been CAT4-tested but is about to be, we recommend almost no preparation. The test is designed to measure reasoning ability rather than knowledge, and over-coaching tends to reduce its usefulness without changing the score much. A short period of familiarity, looking at one or two sample sub-tests so the format is not novel, is sensible. Sleep, breakfast and a relaxed attitude on the day matter more than anything else. We have a related piece on preparing children for admissions interviews that covers the same broad terrain.
Frequently asked questions
What is a good CAT4 score?
A Standard Age Score of 100 is the population average. Above 112 is typically considered high, above 127 is the top 4 percent of the population, and below 88 indicates the child is likely to need additional support in mainstream curriculum work.
Can my child prepare for CAT4?
CAT4 is designed as an aptitude test rather than a knowledge test, and significant coaching tends to reduce its usefulness rather than help the child. A short period of familiarity work to remove novelty effects is sensible. Cramming is not.
Will the school share the CAT4 report with parents?
Most international schools share at least a summary report. Some share the full report with stanines, SAS and group scores. If you would like the detail, ask. Schools generally appreciate parents who want to engage with the data thoughtfully.