Cramer-Von Mises Variance Estimators for Simulations

We study estimators for the variance parameter sigma(2) of a stationary process. The estimators are based on weightings yield estimators that are 'first-order unbiased' for sigma (2) We derive an expression for the asymptotic variance of the new estimators; this expression is then used to obtain the first-order unbiased estimator having the smallest variance among fixed-degree polynomial weighting functions. Although our work is based on asymptotic theory, we present exact and empirical examples to demonstrate the new estimators' small-sample robustness.Naval Postgraduate School, Monterey, California.http://archive.org/details/cramervonmisesva00goldO&MN direct fundingApproved for public release; distribution is unlimited