Test types

The CogAT

A group-administered reasoning test used widely in schools, usually as the first filter for gifted programmes. Understanding the profile codes is most of what a parent needs.

What it is

The Cognitive Abilities Test measures reasoning in three batteries — verbal, quantitative and non-verbal — and is administered to a whole class at once rather than individually.

It is a screening instrument. Group tests are cheap, fast and mechanically scored, which is what makes them practical for testing an entire year group, and the trade-off is that they cannot observe how a child worked or adapt to their level.

Schools typically use CogAT results as a first filter, with individual testing on an instrument such as the WISC reserved for children near a decision boundary.

The three batteries

Verbal. Analogies, sentence completion and verbal classification. Reasoning with words rather than vocabulary knowledge, although the two are hard to separate completely in practice.

Quantitative. Number analogies, number series and number puzzles. Reasoning about quantity, deliberately kept separate from computational skill.

Non-verbal. Figure matrices, paper folding and figure classification. The battery closest to a pure fluid-reasoning measure and the one least affected by language background.

Each battery yields its own score, and the three combine into a composite. The separate scores are usually more informative than the composite, particularly where a child's language background differs from the test language.

How results are reported

CogAT reporting uses several scales at once, which is the main source of parental confusion.

Note the standard deviation: the SAS uses 16, not the 15 used by Wechsler tests. A SAS of 132 is about the 98th percentile, not the 97.7th you would get on a 15-point scale. It is a small difference that matters near a threshold.

Use in gifted identification

Most programmes using CogAT set a threshold on the composite or on a specific battery — commonly the 95th or 98th percentile.

Using a single battery rather than the composite is often the better practice, and some districts do it deliberately. A child with an exceptional non-verbal score and a middling verbal score — a common pattern for children whose home language differs from the test language — is missed entirely by a composite threshold.

As with every threshold, measurement error means a child near the line will fall on either side of it on different days. Multiple-source identification exists precisely because single-test thresholds sort partly on noise.

Preparation

Familiarisation is reasonable: knowing the format, understanding the instructions, having seen a figure matrix before. A child who loses items to unfamiliarity is not being measured accurately, and removing that is fair.

Beyond that the calculation changes. A heavily coached child who gains entry to a programme they are not ready for has a worse year than one who does not, and the coaching does not travel with them into the classroom. Several districts have moved towards less coachable instruments in response to exactly this dynamic.