Table 1.

Classification of Outcomes Among a Family of m Statistical Tests, for m0 of Which the Null Hypothesis Is True




Accept null No discovery Negative result

Reject null Declare discovery Positive result

Total
True null hypothesis U V m0
False null hypothesis T S m1
Total
W
R
m

[i] Column headings give three alternative phrases with the same meaning. The letters R through W represent the number of test that fall into each category. For example, R is the number of tests that are declared significant, rejecting the null hypothesis. The V tests that reject the null when it is true commit type I errors. The T tests that accept the null when it is false make Type II errors.