The belief a calculated tolerance rests on, to annotate the
derivation's collationTolerance with: an inference from the scatter
of the readings, the size of the sample and the rule that fixed the
threshold. used names what the inference worked on, such as the
copies compared.
It is held likely rather than true. The tolerance follows from the
sample only as far as the sample is normal, and a sample far from
that shape makes the number unreliable rather than wrong, which is
what normality is reported for.
The belief a calculated tolerance rests on, to annotate the derivation's
collationTolerancewith: an inference from the scatter of the readings, the size of the sample and the rule that fixed the threshold.usednames what the inference worked on, such as the copies compared.It is held likely rather than true. The tolerance follows from the sample only as far as the sample is normal, and a sample far from that shape makes the number unreliable rather than wrong, which is what
normalityis reported for.