When Does a Score Difference Actually Matter?
Two assessment scores can be different without the difference being meaningful. A statistically significant difference may deserve attention without changing our recommendations. And a small difference may have major consequences when a formal eligibility cutoff is involved.
Responsible interpretation requires understanding which kind of difference we are discussing—and what it means for the person being evaluated.

Different Numbers Do Not Automatically Mean Different Abilities
Consider standard scores of 84 and 85 on a measure with a mean of 100 and a standard deviation of 15. One falls just below one standard deviation from the mean; the other falls exactly at that boundary.
That mathematical distinction does not establish a meaningful difference in functioning. Neither does crossing the boundary between two descriptive categories.
Classification labels help summarize results, but abilities do not abruptly change at the edges of those categories. We should be careful not to describe nearly identical performances as substantially different simply because a label changes.
Statistical Significance Answers a Specific Question
When comparing scores, statistical significance concerns whether the observed difference exceeds what would be expected from measurement error under the comparison method’s assumptions.
We cannot determine that simply by looking at the numbers or their classification labels. We need the appropriate comparison procedures for the measures involved.
Even then, statistical significance is only part of the interpretation. When available, base-rate information adds another question: How frequently does a difference this large occur in the relevant comparison sample?
A statistically significant difference may still be relatively common. It does not automatically establish a disorder, explain its cause, or identify a treatment need.
Practical Meaning Depends on the Question
A difference becomes clinically useful when it helps us understand the person or make a supported decision about their needs.
Does it help explain an everyday difficulty? Does it fit with the history and other findings? Would it change a recommendation?
A small difference can also become consequential because of an external rule. If a program uses a strict score cutoff, one point may affect eligibility. That consequence is real, but it does not make one point a substantial difference in ability.
Our reports should distinguish what the score suggests about functioning from what an institution’s rules require.
Confidence Intervals Help Communicate Precision
An obtained score is an estimate, not a perfectly precise measurement. Confidence intervals help communicate the uncertainty associated with that estimate.
Depending on the measure and score, an interval may extend across descriptive categories or an eligibility threshold. Reporting only the single number can conceal that uncertainty.
Use the interval provided by the appropriate scoring procedures, and explain its relevance in plain language. An interval crossing a cutoff does not automatically establish eligibility; it identifies uncertainty that should be considered alongside the applicable criteria and other evidence.
Give Differences the Weight They Deserve
Good interpretation goes beyond naming score ranges or identifying statistically significant discrepancies. It explains which findings matter, how strongly they are supported, and what they contribute to answering the referral question.
Before emphasizing a difference, ask: Is it statistically supported? How common is it? What does it help explain? Does it change what we recommend?
The numbers deserve careful attention. So does the meaning we assign to them.




Comments