More on Hypothesis Testing for One Population Mean

Remember, to do hypothesis testing for a population mean, you must have:

• Random sample
• Normally distributed data

Hypothesis testing using confidence intervals

If you construct a CI, and the null hypothesis does not lie in the CI, then you reject it. Otherwise, you do not reject it.

If you do it this way, the alternative hypothesis should be two-sided.

The significance level \(\alpha\) should coincide with the confidence level.