Checks whether a small validation sample is consistent with a published or transferred reference interval, via an exact binomial test on the out-of-interval count.
Establishing a reference interval from scratch needs ~120 reference individuals — often impractical. Instead you can adopt a published interval and verify it on a small validation sample (e.g. 20 healthy subjects), confirming that not many more results fall outside than the coverage implies.
With a 95% interval, about 5% of healthy results should land outside by design. The engine counts how many of your validation results fall outside the published limits and runs an exact binomial test of whether that fraction is significantly greater than the expected 1 − coverage. If it isn't, the interval verifies for your population.
A validation-sample column, the published lower and upper limits, and the coverage (e.g. 95%).
The out-of-interval count vs expected, the one-sided binomial p-value, and a verified / not-verified verdict.
With small samples this test has low power — a pass means "no evidence the interval is wrong for your population", not proof it is right. A clear fail (many results outside) is a strong signal to establish your own interval or investigate a population/method difference.