Fill in the blanks

When dealing with a two-factor repeated measures experimental design, S x A x B, where the factor A is fixed and factor B is random, the F-value for the__________________ factor cannot be determined directly. Therefore, we need to use a __________________ F-value, which is obtained by combining the expected values of several mean squares. In order to decide whether the obtained value is statistically significant, the degrees of freedom for the numerator are computed using the_____________ formula. The F-value for the random factor can also not be tested directly (say true or false): .__________________