Bartlett correction of an independence test in a multivariate Poisson model

TL;DR

This paper investigates a Bartlett correction for a likelihood ratio test of independence in a multivariate Poisson model, improving size approximation in low dimensions through explicit correction derivation and simulation analysis.

Contribution

The paper derives an explicit Bartlett correction for the independence test in multivariate Poisson models, enhancing test accuracy in low-dimensional cases.

Findings

Bartlett correction improves test size accuracy in two dimensions

Correction is less effective in higher dimensions

Simulation confirms practical usefulness in small dimensions

Abstract

We consider a system of dependent Poisson variables, where each variable is the sum of an independent variate and a common variate. It is the common variate that creates the dependence. Within this system, a test of independence may be constructed where the null hypothesis is that the common variate is identically zero. In the present paper, we consider the maximum log likelihood ratio test. For this test, it is well-known that the asymptotic distribution of the test statistic is an equal mixture of zero and a chi square distribution with one degree of freedom. We examine a Bartlett correction of this test, in the hope that we will get better approximation of the nominal size for moderately large sample sizes. This correction is explicitly derived, and its usefulness is explored in a simulation study. For practical purposes, the correction is found to be useful in dimension two, but not in higher dimensions.