Our analysis estimated that the share of classrooms where cheating occurred on any particular subject test ranged from 1.1 percent to 2.1 percent, depending on where
the statistical cutoff point was set.
«Given that I know it's safe, and given that I know it's just missing
the statistical cutoff point of being significant, we feel that on an emergency basis, we can distribute it,» Fauci said.
The range in estimates is again explained by the higher and lower
statistical cutoff points we used to determine whether a classroom should be suspected of cheating.
Not exact matches
At first, I assumed it was just a random
statistical artifact of where the annual
cutoff points are.