I'd think if they were only sitting a week or so later, then their results would be included in the main set of results as part of the normal standardised process.
If however they take the test some time after and standardisation has already taken place, then I would think that their result is compared to the rest of the cohort. Statistically there will be very few compared to the main cohort.
However how they deal with the age difference and the fact the test is taken maybe a few months later, I'm not sure. Would they compare with those born in the same month, or would they compare with those born the same number of months older as the number of months later they took the test.
So if a July born doesn't take it in September but in January 4 months later. Are they compared with other July borns or march borns. I'd expect them to be compared with March borns myself.
That would be the fairest way of doing it, grouping all children of the same age (in years and months) at the time they sat the test
From the OP's username and thread title, I infer that their child has sat the Bucks test some time after the standardisation of the main cohort and has not qualified.
OP, if my assumptions are correct and you are considering submitting a review or appeal, focus on collating the academic evidence that suggests suitability for a grammar school place rather than looking for ways that the system may not have worked in your favour. There's lots of good advice on here that will help you through that process.