A psychological experiment was performed to determine whether in problem-solving dyads containing one male and one female, "influencing" behavior depended on the sex of the experimenter. The problem for each dyad was a strategy game called "Rope a Steer," which required 20 separate decisions about which way to proceed toward a defined goal on a game board. For each subject in the dyad, a verbal-influence activity score was derived as a function of the number of statements made by the subject to influence the dyad to move in a particular direction. The difference in verbal-influence activity scores within a dyad was denoted as the variable VIAD, which was then used as the dependent variable in an ANACOVA designed to control for the effects of differing IQ scores of the male and female in each dyad. The relevant data are given in the following tables.
a. State an ANACOVA model appropriate for these data.
b. Determine adjusted mean VIAD scores for each group, and compare these to the unadjusted means for each group.
c. Test whether the adjusted mean scores differ significantly.
d. Find a 95% confidence interval for the true difference in adjusted mean scores.

Regression of Viad on IQM, IQF, and Z

Regression of Viad on IQM, IQF, Z, IQM × Z, and IQF × Z