The report runs to fifteen pages and most of them are tables. You will find your child's name, a column of numbers, and no obvious way to tell whether those numbers are bad news.
Parents read the low ones first. That is human, and it is usually the wrong place to start. Here is what the scales mean, so that the numbers say what they actually say.
The scales you will see
Cognitive tests are scored by comparing your child to a large sample of children the same age. Different tests report that comparison on different scales, and a report will use several at once.
| Scale | Mean | Average range | Where you see it |
|---|---|---|---|
| Standard score | 100 | 85 – 115 | IQ, memory, and academic composites |
| Scaled score | 10 | 8 – 12 | Individual subtests |
| T score | 50 | 40 – 60 | Behavior and attention rating scales |
| Percentile rank | 50th | 16th – 84th | Reported alongside all of the above |
The ranges in that third column are the same range described four ways. Each covers the middle of the distribution — a little over two-thirds of children of that age. A standard score of 92, a scaled score of 9, a T score of 47 and the 30th percentile are all ordinary results.
Average is the most common result on every test in the battery. It is not a consolation prize.
Percentile rank is not a percentage correct
This is the one that causes the most alarm. A percentile rank of 30 does not mean your child got 30% of the items right. It means roughly 30 out of 100 children the same age scored at or below that point — which puts them inside the average range, not near the bottom of it.
Why one low score is not a diagnosis
A full evaluation produces dozens of scores. In any large set of measurements, some will land low simply because that is how distributions work — and that is true of children with no difficulty at all. Research on how often healthy children produce one or more low scores across a battery is the reason a careful clinician does not diagnose from a single number.
What the clinician is doing instead is looking for a pattern: several scores in the same cognitive domain moving together, a profile that matches the history you gave at the consultation, and a result that explains the problem you came in with. One isolated dip in a battery of thirty scores usually explains nothing.
What the report should give you beyond the numbers
- Whether the data supports a diagnosis, and — just as important — where it does not.
- Recommendations specific enough to act on, naming the setting: home, classroom, or clinic.
- Accommodations written in language a school district will recognize.
- What should be re-checked, and when.
Questions worth asking at the feedback session
- Which of these scores actually changed your thinking?
- What would you expect to see in a year if nothing changes? If the recommendations are followed?
- Which recommendation would you do first?
- What did you rule out, and on what basis?
A good report answers most of those before you ask. The session exists so you can ask the rest.
References
- American Educational Research Association, American Psychological Association, and National Council on Measurement in Education. Standards for Educational and Psychological Testing. The reference work governing how these scales are built and reported. The 2014 edition is current and open access; a revision is under way.
- Brooks, B. L., Sherman, E. M. S., and Iverson, G. L. Healthy children get low scores too: prevalence of low scores on the NEPSY-II in preschoolers, children, and adolescents. Archives of Clinical Neuropsychology, 2010;25(3):182–190. Base rates of low scores across a battery in 1,200 healthy children aged 3 to 16.
