A researcher wants to carry out a hypothesis test involving the mean for a sample of size .
She does not know the true value of the population standard deviation, but is reasonably sure that the underlying population is approximately normally distributed. Should she use a z-test or a t-test in carrying out the analysis? Why?