Suppose that a random sample of teenage boys and girls are asked how many Harry Potter books they have read. You conduct a two-sided test of the null hypothesis that μ_girls − μ_boys = 0 and you also calculate a 95% confidence interval for μ_girls − μ_boys.
a. Describe (in words) what the parameter μ_girls − μ_boys means here.
b. Now suppose that your friend analyzes the same data but with the order of subtraction reversed (boys − girls, rather than girls − boys). Describe the impact (if any) on each of the following. In other words, describe how your friend’s findings will compare to yours with regard to each of the following.
i. Distribution of simulated statistics under the null hypothesis
ii. Standard deviation of the simulated statistics under the null hypothesis

iii. Observed value of the statistic (difference in sample means)
iv. Approximate p-value from simulation
v. Value of t-test statistic
vi. p-Value from t-test
vii. Midpoint of confidence interval
viii. Endpoints of confidence interval
ix. Width of confidence interval
c. What is the bottom line: Will you and your friend reach the same conclusions even though you disagreed about the order in which to perform the subtraction (boys − girls or girls − boys)? Explain.