Some approximations to the multivariate hypergeometric distribution with applications to hypothesis testing

In this paper, we will examine some approximations to the multivariate hypergeometric distribution by continuous random variables. The continuous random variables will be chosen so as to have the same range of variation, means, variances and covariances as their discrete counterparts. We then show how these approximations can be used in testing hypotheses about the parameters of the multivariate hypergeometric distribution.