Part
One: Probability
Probability
and Sample Size Requirements
Minimal Sample Size as a
Function of Probability
Suppose
that we have an Event, E, with probability PE. Then the quantity (1/ PE) represents the sample size with an expected count for E of 1. That is,
eE ≈ 1 for n ≈
(1/ PE).
Fixed Sample Size and
Minimum Detectable Probability
For fixed
sample size n, the quantity (1/n) represents the probability value for any
event with a perfect count of 1. That is, if
PE = (1/n) for some event E, then the expected
count for event E in samples of size n is 1. That is,
PE = (1/n)
eE = n*(1/n) = 1 for sample size n.