Tests and scores

Updated 29 October 2026

GL Assessment scores explained

CAT4, the GL Progress Tests and the New Group Reading Test all report results the same way. This guide explains each number on the report, in plain English.

Standard Age Score (SAS): 100 is average for a child of that age. About two in three children score between 85 and 115.

Stanine: a 1 to 9 band. 4 to 6 is average, 7 and 8 above average, 9 (a score of 127 or more) is the top 4%.

National Percentile Rank (NPR): how many children in every 100 of the same age scored the same or lower. 50 is average.

Which reports use these scores

GL Assessment's main school tests all use the same scale: a Standard Age Score with an average of 100 and a standard deviation of 15, plus a stanine and a percentile rank. That includes:

Standard Age Score (SAS)

The SAS is the main number on every GL report (for NGRT, GL calls it "the most important piece of information"). It starts from your child's raw mark, adjusts it for their age in years and months, and places it on a scale that compares them with a nationally representative sample of UK children of the same age. So a child born in August is compared with other children born around the same time, not with classmates nearly a year older.

100 is average. Because the scale has a standard deviation of 15, about 68% of children score between 85 and 115. A score of 115 or 85 is therefore not unusual, while 130 or 70 is.

GL says the SAS is the fairest way to compare children across a year group or across years, and the best number for tracking progress.

Stanines

A stanine (short for "standard nine") groups the scores into nine bands. GL notes they are particularly useful when reporting to parents and students because they are easy to understand. GL's bands:

StanineDescriptionSASPercentileChildren in this band
9Very high127 and above97 and above4%
8Above average119 to 12690 to 967%
7Above average112 to 11878 to 8912%
6Average104 to 11159 to 7717%
5Average97 to 10341 to 5820%
4Average89 to 9623 to 4017%
3Below average82 to 8812 to 2212%
2Below average74 to 815 to 117%
1Very low73 and below4 and below4%

Stanines are deliberately broad. A score of 111 and a score of 104 are both stanine 6, while 111 and 112 sit in different stanines. So use the stanine for the big picture and the SAS for the detail.

Some reports use stanines for parts of a test. NGRT, for example, gives stanines for sentence completion and passage comprehension, and PTE gives stanines for English skills and for reading. GL's rough guide for NGRT is that a gap of two or more stanines between parts may show a real difference worth looking into.

National Percentile Rank (NPR)

The NPR is the percentage of children of the same age in the national sample who scored the same or lower. An NPR of 50 is average. GL's examples: an NPR of 5 means the score is within the lowest 5% of the sample, and 95 means it is within the highest 5%.

Percentiles stretch out in the middle and squeeze together at the ends. Moving from 100 to 105 lifts the percentile from about 50 to 63; moving from 130 to 135 only lifts it from about 98 to 99.

Score lookup

SASStaninePercentile (about)Description
140999Very high
135999Very high
130998Very high
127996Very high
125895Above average
120891Above average
119890Above average
115784Above average
112779Above average
110675Average
105663Average
104661Average
100550Exactly average
97542Average
95437Average
90425Average
89423Average
85316Below average
82312Below average
8029Below average
7525Below average
7424Below average
7012Very low

Percentiles are worked out from the SAS scale (average 100, standard deviation 15). The NPR printed on your child's report comes from GL's own tables for that test and can differ by a point or two. For every CAT4 score from 69 to 141, see CAT4 scores explained.

Other numbers you may see

Group rank

Your child's position within the group the school tested, such as the class or year group. An equals sign means a shared rank. It says nothing about the national picture.

Progress category

On the Progress Tests, if your child has taken the test before, the report compares the two results and gives one of five categories: much higher than expected, higher than expected, expected, lower than expected or much lower than expected. NGRT's progress report flags higher and lower than expected progress in a similar way. "Expected" means progress in line with what the earlier score would predict, which is a perfectly good result.

Reading age

NGRT also gives a reading age: the age at which a score is most likely to be achieved. GL warns it is not sensible to use reading ages for above-average readers and recommends the SAS for tracking progress. See NGRT scores and reading age.

Indicators

Teacher reports for CAT4, and for the Progress Tests, can include indicators of likely results in later UK exams. See CAT4 GCSE indicators and predicted grades.

Every score has a margin of error

No test measures perfectly. GL's technical documents give a 90% confidence band for each test: for an average child, there is a 90% chance the "true" score lies within this many points of the reported score.

Test90% band (+/- points)
CAT4 (mean of the four batteries)6
CAT4 single battery (verbal, quantitative, non-verbal, spatial)8 to 10
Progress Test in Maths (PTM 10)6
Progress Test in English (PTE 12)7
Progress Test in Science (PTS 10)9
New Group Reading Test (NGRT)8

GL also says that differences of less than 3 points are not usually statistically significant. So a child who scores 112 one year and 109 the next has, in practice, held steady.

Scores for children at schools in the UAE

GL uses its UK standardisation for international schools, so 100 always means the average for a UK child of the same age. GL has compared international schools with that average:

  • CAT4: in GL's study of about 224,000 international students, the average was about 102 overall, slightly higher in quantitative, non-verbal and spatial reasoning but lower (about 97) in verbal reasoning. GL links this to the number of children learning English as an additional language.
  • Progress Tests: international schools scored above 100 at every level in GL's 2023 benchmark, most of all in science.

The practical point: a score of 100 is exactly average nationally, but it may be below the middle of your child's class. Compare your child with their own earlier scores first.

Questions parents ask

What does standard age score mean?

A Standard Age Score (SAS) is a test mark that has been adjusted for the child's exact age and placed on a scale where 100 is average for that age. GL Assessment uses it on CAT4, the Progress Tests and NGRT. The scale has a standard deviation of 15, so about two in three children score between 85 and 115.

What does stanine 9 mean?

Stanine 9 is the top band of the nine. In GL's bands it covers a Standard Age Score of 127 or more, which is the top 4% of children of the same age (97th percentile and above). GL describes it as 'very high'.

What percentile is stanine 9?

The 97th percentile and above. Stanine 8 is the 90th to 96th percentile, and stanine 7 the 78th to 89th.

What does a stanine of 5 mean?

Exactly the middle band: Standard Age Scores of 97 to 103, the 41st to 58th percentile. One child in five lands here. Stanines 4, 5 and 6 together are the 'average' range and cover more than half of all children.

What does National Percentile Rank mean?

The National Percentile Rank (NPR) tells you how many children in every 100 of the same age in the national sample scored the same as or lower than your child. An NPR of 50 is average; an NPR of 90 means your child scored as well as or better than 90 in every 100.

Is the standard age score the same in the UK and the UAE?

Yes. GL Assessment uses its UK standardisation for schools outside the UK, so 100 means average for a child of that age in the UK sample. Because international schools tend to score above the UK average, the average in your child's class may be higher than 100.

Why has my child's score gone down?

First check the size of the change. GL says differences of under 3 points are not usually significant, and every score has a margin of error of roughly 6 to 10 points either side. A drop larger than that is worth discussing with the teacher, together with the progress category if the report shows one.

Can I compare my child's CAT4 score with their Progress Test score?

They use the same scale, but measure different things: CAT4 measures reasoning, the Progress Tests measure what has been learned. GL's combination report puts them side by side and flags children whose attainment is higher or lower than their CAT4 scores would suggest.

Sources