Question - A recent study investigated the association of well-being with various health risk behaviors in college students. The sample consisted of 9515 undergraduate students, 6945 females and 2570 males. The following table presents the sample proportions, by gender, of various risky behaviors performed in the last 30 days.
(a) For each of these behaviors, test whether there is a gender difference in the proportions, using the α = 0.05 significance level.
(b) Since, for each behavior, the probability of a Type I error is 0.05, the probability of at least one Type I error is greater than 0.05. Let's adjust for this using the Bonferroni procedure. Suppose that there are a total of 10 behaviors so that the significance level for each test should be 0.05/10 = 0.005. Does this change in significance level alter any of your conclusions?
(c) Suppose that differences in proportions greater than 0.06 are considered of practical significance. For which risk factors do you find both practical and statistical significance?