14.1 Consider the following data set:

B1 B2 B3 S11 23 16 12 A1 S21 27 1 14 S31 22 10 10 S12 25 16 19 A2 S22 17 33 22 S32 9 17 22

(a) Present the complete ANOVA table with all numerical results; assume A and B both have fixed-effects.

(b) Find the mean score for each subject. You now have a one-factor completely randomized design. Perform an ANOVA using the mean scores as the data. (i) How does the F test of A compare with that calculated in part (a)? (ii) How does MSA in this analysis compare with that obtained in part (a)? Explain the relation.

(c) In this chapter, we stated that SSSB/A was equivalent to the result of calculating the S×B sum of squares separately at each level of A and then summing the a terms. To demonstrate this, ignore the A2 data and calculate the sum of squares for S×B at A1 Do the same