Chapter 5 illustrated that when (randomly) sampling observations from a skewed distribution where outliers are rare, it is generally reasonable to assume that plots of sample means over many studies has, approximately, a normal distribution. This means that reasonably accurate confidence intervals for the mean can be computed when the standard deviation, σ, is known. Based on information summarized in this section, how does this contrast with the situation where the variance is not known and confidence intervals are computed using Student’s t-distribution?