A standard score places a result on a fixed common scale so that findings from different areas can be compared on the same footing. It is not a mark out of a hundred, and it isn't a percentage of anything. The scale has a set average and a set spread, and a standard score tells you where a result sits on it.
I'm not a clinician, and it took a psychologist explaining this twice before it clicked. I'm writing down the version that worked, because it's a statistics question rather than a clinical one — which makes it one of the few things on an assessment report I'm actually qualified to help with. What any particular score means for a particular person belongs to the clinician who did the assessment.
A common scale exists so that results from different areas can be compared fairly.
What is a standard score, in plain terms?
It's a translation.
Different tasks produce raw results that aren't comparable — one might be out of 40 items, another timed, another scored by a rubric. You can't line those up meaningfully. So each raw result is converted onto a common scale defined by two things: an average and a spread.
Once everything is on that scale, a result in one area can be compared fairly with a result in another. That's the entire purpose. It's a unit conversion, not a grade.
The explanation that finally landed for me was this: a standard score answers "how far from typical, and in which direction?" — nothing more. Whether being that far from typical matters is a separate question, and it isn't a statistical one.
Why isn't the scale out of 100?
Because the scale isn't measuring completion, it's measuring position, and there's no natural maximum.
You'll see more than one scale on a single report, which trips people up. Some areas are reported on a scale with an average around 100; sub-areas are often reported on smaller scales with a different average. Both are doing the same job. Neither is a percentage.
This is worth knowing because the most common misreading I hear about is someone seeing a number below 100 and reading it as a failing mark. It isn't a mark at all. A result somewhat below the average of a scale is, by definition, where a large number of people sit.
How does a standard score relate to a percentile?
They're two views of the same position.
A standard score expresses the position in units of the scale. A percentile expresses the same position as "at or above roughly this share of the comparison group". A report often gives both, and they should agree — if they seem not to, that's a question for the clinician, not a contradiction you need to resolve.
The reason both appear is that they're useful to different readers. The scale version travels well between areas and between reports; the percentile version is easier to picture. There's a fuller explanation of the second one in what 'percentile' means on an assessment report.
A standard score and a percentile are two views of the same position.
Why are results described in bands rather than exact points?
Because precision beyond a certain point would be false, and psychologists are careful about that.
Most reports group results into descriptive bands — a broad middle region covering a large share of people, with regions above and below. Two results that differ numerically can land in the same band, and when they do the report will usually treat them as equivalent. That isn't hedging. It's the document declining to make a distinction the measurement can't support.
You'll often also see a bracketed range after a score. That's the report acknowledging that any single measurement is a sample taken on one day, in one room, with a person who might have been tired or unusually sharp. More than one psychologist told me they'd rather families read the range than the point estimate, and include it for exactly that reason.
What can't a standard score tell you?
The part that matters most, which is why the report has all those paragraphs around the numbers.
It can't tell you why a result came out where it did — whether something hasn't been taught, hasn't been practised, or is genuinely effortful. It can't tell you what to do next. It can't be read on its own, because the shape across areas usually carries more meaning than any single point in it. And it isn't a fixed property of a person; it's a measurement of performance on a particular set of tasks at a particular time.
Turning those numbers into something a person can act on is the clinician's work, and it's the part of the document that takes the longest — there's more on how that gets built in from raw scores to report narrative, and the surrounding structure is in what's in a cognitive assessment report.
If a number in your report is worrying you, ask the person who wrote it. Every psychologist I've put that to has said the same thing: they would much rather explain it again than have it sit on a page being misread.
For clinicians reading this — I wrote it as something you can send home, because the same question kept coming back to you. The write-up itself is the part our drafting is built to give back.
— Ian
