Order Statistics and Parameter Estimation in the Generalized Lambda Distribution

Abstract

SYNOPTIC ABSTRACTIn this paper, we first derive the exact forms for single and product moments of order statistics from the generalized lambda distribution. These forms are written in terms of beta and hyper geometric functions and used to obtain the necessary coefficients for the best linear estimators of the location and scale parameters. We present three examples to illustrate the importance of the moments and variance-covariance structure in obtaining the best linear estimators and compare these estimators with the estimators obtained by other procedures.