Explain what a 90% confidence interval for the population mean is. Why must we assume that the null hypothesis is true in order to compute a p-value when doing a hypothesis test for the mean of a normally distributed population? How does the sample size affect the margin of error in a confidence interval?