Exact expressions for the expected value, variances and covariances of the order statistics from a random sample from the Cauchy distribution are derived in terms of infinite series. An error analysis deremines the number of terms in these convergent series needed to attain any desired accuracy, establishing their superiority to the numerical integration methods used by Barnett 1966a. A comparison of explicit estimation methods for the local and scale parameters in the general Cauchy distribution establishes that the Tiku-Suresh 1992 technique is the most advantageous
展开▼
