Published in Romanian Journal of Economic Forecasting, volume 7 issue 2, 2006.
| Requires a PDF viewer such as Xpdf
In a previous study, (Stefanescu, P., Stefanescu, St., 2006) we suggested two
possible estimators for the unknown value of the parameter q which characterize
a homographic type HG(q) distribution.
In the current paper we shall prove that the proposed estimators of q based on the median of the r.v. X, X ~ HG(q) , always over- evaluate the real value of q . For this reason we determined an adjusted multiplicative factor such that the median type estimators to become unbiased.
The theoretical results were confirmed experimentally by using a Monte Carlo stochastic simulation technique.
homographic distribution, unbiased estimator, Monte Carlo technique, generating
JEL Classification: C13, C15