I currently have a text book problem that I cant seem to work out:
Suppose X1,...,Xn~Exp(c) and consider the estimate of the variance sigmahat2(X1,...,Xn)=1/n [n][/i=1] (Xi-Xbar)2. Assume c is known.
For c=2, n=10 determine a Monte Carlo (MC) estimate of the bias of sigmahat2. Carefully choose the MC sample size N such that the error of the estimate is small compared to the result.
Also, write an algorithm for given c and n that first determines an appropriate N for the estimate of the bias of sigmahat2 and then uses this to compute a good estimate of the bias.
Any help would be greatly appreciated!