A reputable medical journal reported on an experiment intended to see if a certain drug could be used as a treatment for the eating disorder anorexia nervosa. The subjects, women being treated for anorexia, were randomly divided into two groups. Of the 49 who received the drug, 35 were deemed healthy a year later, compared to 31 of the 44 who got the placebo. The assumptions and conditions for inference are satisfied and the 99% confidence interval for the difference in outcomes, where the women who received the drug are sample 1 and the women who received the placebo are sample 2, is (-0.233,0.253). Suppose that a hypothesis test had been conducted instead. Complete parts a through e below without actually doing the test. a) What hypotheses would be tested? Ho P1 P2 HA P1 P2 19 (Type integers or decimals. Do not round.) b) State a conclusion based on the confidence interval. Ho. There sufficient evidence to conclude the difference between the proportion of anorexic women who were healthy after taking the drug and the proportion of anorexic women who were healthy after taking the placebo is (Type an integer or a decimal. Do not round.) c) What alpha level did this test use? (Type an integer or a decimal. Do not round.) d) If the part b conclusion is wrong, which type of error was made? If the part b conclusion is wrong, it means that the test the null hypothesis when it was actually This is a Type error. e) What would be the consequences of such an error? A. A drug that helps women overcome anorexia nervosa would likely be discarded instead of sold to the public. B. A placebo that does not help women with anorexia nervosa would likely be marketed and sold to the public as a treatment for anorexia nervosa. C. A drug that does not help women with anorexia nervosa would likely be marketed and sold to the public as a treatment for anorexia nervosa. D. The drug would need to be retested for effectiveness.