Type I or Type II?

The following table has helped me avoid confusing Type I and Type II errors. Write the true hypothesis numbers on the left, and the ones announced on top.

	Announced 0	Announced 1
Actually 0	No error	Type I
Actually 1	Type II	No error

Now read out the row and column headers as a binary number:

• 01 (which is 1 in binary) means H0 was true and we announced H1. That is error Type I.

• 10 (which is 2 in binary) means H1 was true and we announced H0. That is error Type II.

(Or, perhaps, it is easier to remember that Type I is a false alarm and Type II is a miss.)