A percentile on an assessment report describes where a score sits compared with other people of the same age. A percentile of 40 means the score is at or above roughly 40 percent of that comparison group. It is not a percentage of questions answered correctly — and that single misreading causes more unnecessary worry than anything else on the page.
I'm not a clinician, and I had to have this explained to me properly. I'm writing it down because psychologists told me it's the question that comes back on report after report, and because it's a statistics question rather than a clinical one — which makes it one of the few things on a report I can actually help with. What any particular percentile means for a particular person is a question for the clinician who did the assessment.
A percentile is a position in a group, not a mark out of a hundred.
What does a percentile actually mean?
It's a position, not a score.
Imagine a hundred people of the same age lined up by result on one task, lowest to highest. A percentile of 40 puts the person at roughly the fortieth position from the bottom — at or above about 40 percent of that line-up. The number is entirely about where in the group, and says nothing on its own about how many items were answered, how hard they were, or how the person felt doing them.
Two consequences worth holding onto. The comparison group matters — results are compared against people of a similar age, so the same raw performance produces a different percentile at different ages. And percentiles are not evenly spaced: near the middle of the distribution, large numbers of people are packed into a narrow band of actual performance, so a move from the 45th to the 55th percentile can represent a very small real difference.
Is a percentile the same as a percentage?
No, and this is the confusion worth clearing first because it's the one that frightens people.
A percentage is out of the questions asked: 40 percent means 40 of 100 items correct. A percentile is out of the people: the 40th percentile means the result is around or above 40 percent of the comparison group.
So a percentile of 40 is not "a failing grade". It's a position somewhere in the middle of a normal spread of people. Nothing about the number by itself indicates a problem — which is exactly why clinicians write interpretation around it rather than leaving numbers to speak for themselves.
Why is "average" a range instead of a single point?
Because with any characteristic that varies across people, most people cluster in the middle, and drawing a line at one exact point would be arbitrary.
Reports therefore describe bands rather than a single average — a broad middle region covering a large share of people, with regions above and below. When a report says a result falls in an average range, it's saying the result sits in the same territory as a large proportion of the comparison group, not that it hit a target.
This is also why two results that look different numerically can be described in the same band. That isn't the report hedging. It's the report being accurate about how much difference the numbers actually represent.
Most people cluster in the middle, which is why reports describe bands rather than a single point.
Why do psychologists give a range after the number?
Because no single measurement is exact, and the range is the honest version of the result.
Any assessment result is a sample of behaviour on a particular day, in a particular room, with a particular person, when someone might have been tired or nervous or unusually focused. The bracketed range you often see after a score — the confidence interval — is the report saying: our best estimate is this, and the true value is very likely to sit somewhere in this band.
More than one psychologist told me they'd rather families read the range than the point estimate, and that they include it precisely so a single number doesn't get treated as a fixed fact about a person. If your report has ranges in it, that's the clinician being careful, not being vague.
What can't a percentile tell you?
Quite a lot, and this is the part I'd most want a reader to take away.
It can't tell you why a result came out where it did. It can't distinguish between a skill that hasn't been taught and one that's hard to acquire. It can't tell you what will happen next, or what should be done about it. And it can't be read in isolation from the rest of the profile — the shape across areas usually carries more meaning than any single position in it. That synthesis is what the integration section of the report is for, and it's the part that requires professional judgment; there's more on how it gets built in from raw scores to report narrative.
If a number in your report worries you, that's a completely reasonable thing to raise with the clinician who wrote it. Every psychologist I've asked has said the same thing: they would far rather explain it again than have it sit on a page being misread. And if you want the wider map of the document those numbers live in, that's what's in a cognitive assessment report.
For the clinicians reading this — I wrote it because the same question kept coming back to you, and a plain page you can send home is cheaper than answering it forty times. The write-up itself is the part our drafting is built to give back — what any number means for the person in front of you stays entirely your call.
— Ian
