# On the asymptotic error of a bivariate normal approximation with an application to simple random sampling

This paper gives a formula for the limiting error of the central limit theorem for the bivariate case. Insight this type of error simplifes the proofs of central limit theorems in probability sampling from finite populations.