Hi all In CT3-12 we test H0, and if it fails at a certain level we say that H0 is rejected with (for example) 95% confidence and H1 is accepted. But since we have not tested H1, how do we know that H1 is better? Maybe both hypotheses should be rejected at that level! This should be so especially where H1 is simple, e.g. in Question 12.1 or in the example given in the Core Reading in the middle of page 8, since then H1 may really be worse than H0! Thank you
Yeah - simple H0 and simple H1 are not real life. Read life tends to be \(H_0: \mu = 5 vs H_1: \mu \neq 5 \) You only get one side confidence intervals if you have some theory as to why it should not be two-sided.
