The likelihood ratio approximation to the conditional distribution of the maximum likelihood estimator in the discrete case

Authors
Citation
Ta. Severini, The likelihood ratio approximation to the conditional distribution of the maximum likelihood estimator in the discrete case, BIOMETRIKA, 87(4), 2000, pp. 939-945
Citations number
18
Categorie Soggetti
Biology,Multidisciplinary,Mathematics
Journal title
BIOMETRIKA
ISSN journal
00063444 → ACNP
Volume
87
Issue
4
Year of publication
2000
Pages
939 - 945
Database
ISI
SICI code
0006-3444(200012)87:4<939:TLRATT>2.0.ZU;2-F
Abstract
The likelihood ratio approximation, also called Barndorff-Nielsen's approxi mation and often denoted by p*, provides a highly accurate approximation to the conditional density of a maximum likelihood estimator <(<theta>)over c ap> given an ancillary statistic. In this paper, the properties of p* are c onsidered for the case in which the underlying random variables have a latt ice distribution and <(<theta>)over cap> has a discrete, but not necessaril y lattice, distribution. If <(<theta>)over cap> has a lattice distribution, ;then p* provides a valid approximation to the density of <(<theta>)over ca p> with respect to counting measure. If the distribution of <(<theta>)over cap> is non-lattice, then p* is still a valid approximation to the conditio nal density of <(<theta>)over cap> however, the dominating measure is no lo nger counting measure.