From the ANOVA table, MSE = 21.722.

F = 124.812/21.722 = 5.746.

The F of 5.746 = t² = 2.397².

The probability value for F (as determined from an F table using dfn = 1 and dfe = 18) is 0.028, which is the same as for t.

Scheffé's test is used for unplanned comparisons. This test is the same as the F test for planned comparisons just discussed except that dfn = a-1. Therefore,

MSB = SSB/dfn = SSB/(a-1)

As in the section on planned comparisons, MSE is from the analysis of variance. It equals: SSE/dfe.