Suppose we have a dataset of exam scores:
Because sample variance uses degrees of freedom. One degree of freedom is "used up" to calculate the mean. Dividing by ( n-1 ) makes ( s^2 ) an unbiased estimator of the true population variance.
