The classical method used to construct confidence intervals for the variance of a normal population does not always yield the
best coverage accuracy. In fact
Contrary to the classical method that uses the Chi-square statistic, the proposed approach relies only on the observations. With the proposed approach, approximate Bayesian confidence intervals for a normal population variance are easily computed for any level of significance.
The approximate Bayesian approach under to the popular square error loss function does not always yield the best approximate Bayesian results. In fact, the Higgins-Tsokos loss function performs better in the above examples.