Skip to content

CogAT vs NNAT vs OLSAT vs CCAT: Which Ability Test Is Your Child Taking?

Last Updated on 18 November 2026 by Editorial Team

All four measure reasoning rather than what a child has been taught, all four report a score with a mean of 100, and three of the four use a standard deviation of 16 rather than the 15 most people assume.

That last detail is not pedantry. A standard deviation of 15 puts one deviation above the mean at 115; a standard deviation of 16 puts it at 116, and the gap widens further out. At the two deviation mark used by many gifted programmes, the difference is 130 against 132, which is the difference between qualifying and not for a real number of children every year.

The four tests compared

The four tests at a glance
Test Publisher Headline score Scale What it emphasises
CogAT Riverside Insights Standard Age Score (SAS) Mean 100, SD 16 Verbal, quantitative and nonverbal batteries, reported separately and together
NNAT3 Pearson Naglieri Ability Index (NAI) Mean 100, SD 15 Nonverbal reasoning only, using shapes and patterns
OLSAT Pearson School Ability Index (SAI) Mean 100, SD 16 Verbal and nonverbal reasoning, weighted toward verbal at older ages
CCAT-7 Nelson Standard Age Score (SAS) Mean 100, SD 16 The Canadian counterpart of the CogAT, same three battery structure

The practical distinction is what each one asks a child to do. The NNAT is entirely nonverbal, which is why districts with many English learners favour it. The OLSAT leans verbal as children get older. The CogAT and CCAT-7 report three separate batteries, which produces a profile rather than a single number and is more informative when the batteries disagree.

Why the same child scores differently on each

Parents often see a child score at the 95th percentile on one and the 80th on another and conclude one of them is wrong. Usually neither is.

These tests sample different abilities. A verbally strong child with weaker spatial reasoning will score higher on the OLSAT than the NNAT, and the gap can be substantial. A child who reasons well with patterns but reads slowly will show the reverse. That is the tests working correctly, not failing.

The second reason is the norm group. Each publisher norms against its own sample, and those samples are not identical. A percentile is always a comparison against a specific group of children, which is the point percentiles, stanines and grade equivalents explained makes at length, because it affects every score report a child ever brings home.

Stanines, which all four report

Stanines, which all four tests report alongside percentiles
Stanine Percentile range Description
1 1 to 4 Well below average
2 5 to 11 Below average
3 12 to 23 Below average
4 24 to 40 Average
5 41 to 60 Average
6 61 to 77 Average
7 78 to 89 Above average
8 90 to 96 Above average
9 97 to 100 Well above average

The theoretical stanine distribution, derived from the normal curve. Individual publishers round the boundaries slightly differently, so a percentile on a band edge can map to either stanine depending on whose table is used.

Stanines compress a percentile into a single digit from 1 to 9, which makes them easy to read and lossy. A child at the 77th percentile and one at the 60th both sit in stanine 6, and the difference between them is real. When a report gives both, read the percentile.

They also differ slightly between publishers, because the bottom band has to absorb the fact that percentile ranks are usually reported 1 to 99 rather than 0 to 100. That is a one point difference at the edge of a band, and it is the sort of thing that decides a borderline case.

What a CogAT profile tells you that a single score does not

The CogAT and CCAT-7 report verbal, quantitative and nonverbal batteries separately, and the pattern between them is called the ability profile. A child scoring evenly across all three has a different educational need from one with a twenty point gap between verbal and nonverbal, even if their composite scores match exactly.

Schools that use the profile well group children by the pattern rather than the total. Schools that ignore it are discarding the most useful thing the test produces. Our CogAT score chart sets out the batteries and the profile codes, and the CCAT-7 page covers the Canadian equivalent.

How these differ from achievement tests

This is the distinction that matters most and gets confused most often. Ability tests ask how a child reasons. Achievement tests ask what a child has learned. They answer different questions and are used for different decisions.

The i-Ready Diagnostic and NWEA MAP Growth are achievement tests: they measure progress against grade level content, which is why they are given two or three times a year and why growth is the useful output. The comparison between those two goes through the difference in detail.

An ability test is usually given once, because reasoning does not change much between October and March. A child can be above average in ability and below grade level in achievement, and that combination is precisely what gifted identification is supposed to catch.

The Woodcock Johnson IV sits across both camps, with separate achievement and cognitive batteries, which is why schools sometimes use it when the picture from a group test is unclear.

If your child is near a cutoff

Most gifted programmes set a threshold, commonly the 95th or 98th percentile, sometimes on a composite and sometimes on any one battery. The exact rule varies by district and is worth asking about directly, because it changes what a borderline score means.

Two things are worth knowing. A single test result carries measurement error, and a child two points below a cutoff is statistically indistinguishable from one two points above it. And many districts will consider a second test or a portfolio, particularly where a child’s profile is uneven.

What rarely helps is coaching. Reasoning tests are reasonably resistant to it, familiarity with the format is worth a modest amount on the first sitting and little thereafter, and a score inflated by drilling sets a child up for a programme that does not suit them.

Common questions

What is a good CogAT score?

The Standard Age Score has a mean of 100 and a standard deviation of 16, so around 116 is one deviation above average and roughly the 84th percentile. Most gifted programmes look for the 95th percentile or above, though the rule varies by district.

Is the NNAT easier than the CogAT?

Neither is easier in general. The NNAT is entirely nonverbal, so children with strong spatial reasoning or limited English often score better on it, while verbally strong children usually do better on the CogAT or OLSAT.

Why is the standard deviation 16 rather than 15?

It is a publisher convention and three of these four use 16. It matters at the thresholds: two deviations above the mean is 130 on a 15 point scale and 132 on a 16 point one.

What is a stanine?

A compression of the percentile into a single digit from 1 to 9. Stanine 5 is the middle, covering the 41st to 60th percentile. It is easy to read and discards detail, so prefer the percentile when both are given.

Can my child prepare for an ability test?

Familiarity with the question formats helps a little on a first sitting. Beyond that these tests are fairly resistant to coaching, and an inflated score is not in a child’s interest if it places them in a programme that does not fit.

Which test do most gifted programmes use?

It varies by district. CogAT and NNAT are the most widely used in the United States, OLSAT is common in some large districts, and CCAT-7 is the Canadian equivalent of the CogAT.

Sources

Riverside Insights. CogAT Form 7 and 8 score interpretation materials, for the Standard Age Score scale and ability profiles.
Pearson. NNAT3 and OLSAT technical information, for the Naglieri Ability Index and School Ability Index scales.

Leave a Reply

Your email address will not be published. Required fields are marked *