Another sampling distribution Use the Normal population at the top. Every statistic has its own sampling distribution. Let’s examine the sampling distribution of the sample variance. Choose N = 5 and Variance at one of the levels below. The sample variance calculated uses the denominator n rather than the n – 1 we are accustomed to using. Is the centre (mean) of the sampling distribution of this sample variance equal to the actual population variance (printed out at the top next to the population)? If so, we say that this statistic is an unbiased estimator of the population variance. Try changing the statistic to Var (U). This statistic uses n – 1 as its denominator. Which version of the sample variance do you prefer? Explain. (We’ll learn more about this denominator, which we call the degrees of freedom, in Chapter 18.)