Example

Aa x aa testcross; x = y = 1/2, n = 8

If we test Ho:1Aa: 1aa and set α = 0.05, then we could only reject Ho when we have 8 Aa or 8 aa progeny. If we set a = 0.10, then we would reject Ho when we have either 8 Aa, 7 Aa, 1 Aa, or 0 Aa in a family of size 8. When we reject Ho this means that our observed data does not support this hypothesis. We realize that sometimes we erroneously reject Ho when it is actually true, due to an unusual sample of data.