The Women’s Health Initiative conducted a randomized experiment to see if hormone therapy was helpful for postmenopausal women. The women were randomly assigned to receive the estrogen plus progestin hormone therapy or a placebo. After five years, 107 of the 8506 on the hormone therapy developed cancer and 88 of the 8102 in the placebo group developed cancer. Is this a significant difference?
a. Set up notation, and state assumptions and hypotheses.
b. Find the test statistic and P-value, and interpret. (If you prefer, use software, such as MINITAB, for which you can conduct the analysis by entering summary counts.)
c. What is your conclusion, for a significance level of 0.05? (The study was planned to be eight years long but was stopped after five years because of increased heart and cancer problems for the therapy group. This shows a benefit of doing two-sided tests, as results sometimes have the opposite direction from the expected one.)