Hi, When calculating method of moments, the IAI manual divides the var(X) by a factor of n. With regards to this, wont using 1/(n-1) be better as that is an unbiased estimator of variance?
Strictly, dividing by n is correct if you have the whole population, and n-1 if you are working with a sample. In practice, the uncertainty in the estimate dominates the difference that is due to n or n-1.
