In September 1996, U.S. News & World Report published a report on America's health maintenance organizations (HMOs). The report was intended to serve as a consumer guide to HMO quality. For each HMO included in the report, data were provided on several variables, including the following:
PHYSTURN: Physician turnover rate (%).
PHYSCERT: Percentage of doctors who were board certified.
PREV: Prevention score, indicating how well the HMO meets Public Health Service goals in various measures of preventive care (including immunizations and prenatal care). The results can be negative (indicating that the HMO falls short of the goals) or positive (indicating that it exceeds the goals).

Question continues to the next page…

a. In the U.S. News & World Report story, the average physician turnover rate for HMOs is reported to be 6%. Suppose that an ANOVA is to be performed to compare the average prevention scores for HMOs with low physician turnover rates (≤6%) and HMOs with higher rates (>6%). State precisely the ANOVA model.
b. In the above SAS output, complete the ANOVA table. To produce the output, the following coding scheme was followed: TURN = 1 if PHYSTURN ≤ 6%; 2 otherwise.
c. Test whether the average prevention scores differ significantly for the two types of HMOs mentioned in part (a).
d. Suppose that an ANOVA is to be performed to compare the average prevention scores for HMOs with low percentages of board-certified primary care physicians (≤75%) and HMOs with higher percentages of board-certified physicians (>75%). State precisely the ANOVA model.
e. In the second part of the above SAS output, complete the ANOVA table. To produce the output, the following coding scheme was followed: CERT = 1 if PHYSCERT ≤ 75%; 2 otherwise.
f. Test whether the average prevention scores differ significantly for the two types of HMOs mentioned in part (d).