What Does a Confidence Interval Mean in an Assessment Report?
Updated: Oct 28
Why the report gives a range as well as a number
A child may receive a standard score of 82 on a particular assessment task, with a confidence interval shown beside it. This can look as though the report is giving several different results. In fact, the score is an estimate from a standardised test, while the interval expresses uncertainty around that estimate.
All tests have some measurement error: a child's performance can be influenced by the particular items, the conditions and ordinary day-to-day variation. A careful report acknowledges this rather than presenting every number as perfectly exact.
What does 95% confidence mean?
In simplified terms, a 95% confidence interval is produced by a statistical method designed so that, across many comparable samples or administrations under the model, about 95% of the intervals would contain the underlying score being estimated.
It does not mean there is a 95% chance that this particular child's true score lies inside the displayed range. That distinction is technical, but the practical message is straightforward: the reported score has uncertainty, and the assessor should avoid treating tiny numerical differences as decisive.
An example of how to read the interval
Suppose a hypothetical report gives a spelling standard score of 82 with a 95% confidence interval of 76–90. The precise interval here is only an illustration, not a prediction for any particular test. It would be unwise to say that the child has an exact and unchanging spelling level of 82.
Equally, the interval does not erase the evidence of spelling difficulty. The assessor should combine the result with the child's spelling in schoolwork, developmental history, other test findings and the nature of their errors.
Do overlapping intervals mean scores are the same?
Not necessarily. Parents may see a reading score and a spelling score whose intervals overlap and assume there is no meaningful difference. Or they may see non-overlapping intervals and assume a particular diagnosis is proved. Neither shortcut is reliable.
Tests may have different measurement properties, and comparing two scores properly may require specific statistical procedures described in the manuals. The assessor should explain any important differences using appropriate evidence, not simply by looking at whether two printed ranges overlap.
Why intervals may be different widths
Some tasks estimate performance more precisely than others. Reliability, test design, age group and the scoring method can affect the width of an interval. A wider interval means more uncertainty around the estimate, not necessarily that the child was inconsistent or uncooperative.
Where results are close to a descriptive boundary, a report should be particularly cautious about using the boundary as if it were a hard dividing line. One or two score points may change a label without changing what the child needs in lessons.
Confidence intervals do not diagnose dyslexia
Dyslexia identification involves a reasoned synthesis of evidence about reading and spelling, relevant processing skills, developmental history and educational impact. A low score with a confidence interval is one piece of that evidence.
It does not automatically establish a diagnosis; nor does an interval reaching into the average range automatically rule out a persistent difficulty. Consider a child who has accurate but exceptionally effortful reading: timed fluency, classroom demands and background information may be essential to understanding the impact.
What should you ask the assessor?
Ask what the test measures, why that particular interval was reported and whether the result is consistent with other evidence. If the report highlights a difference between two scores, ask whether it is statistically and practically meaningful.
If you are deciding what to do at school, move beyond the numbers: which tasks are difficult, what teaching is needed and how will progress be reviewed? An interval should encourage careful interpretation, not create anxiety about an apparently exact figure.
A parent-friendly way to explain it
You can think of a standard score as the best estimate from that assessment and the confidence interval as a reminder that the estimate is not perfectly precise.
It is similar to giving a measurement with a margin of uncertainty, although the statistics behind it are more specific. The most useful report translates that uncertainty into a clear description of the child's learning profile and sensible recommendations.