Sharp exponential inequalities in survey sampling: conditional Poisson sampling schemes

TL;DR

This paper develops exponential deviation bounds for sample sums in rejective survey sampling schemes, overcoming dependence issues by combining Poisson sampling formulation with the Escher transformation, leading to more accurate inequalities.

Contribution

It introduces Bennett/Bernstein type bounds for rejective sampling, extending classical inequalities to dependent sampling schemes using a novel approach.

Findings

Derived exponential bounds for rejective sampling deviations.

Showed bounds are more accurate than negative association-based inequalities.

Extended results to other sampling schemes approximable by rejective plans.

Abstract

This paper is devoted to establishing exponential bounds for the probabilities of deviation of a sample sum from its expectation, when the variables involved in the summation are obtained by sampling in a finite population according to a rejective scheme, generalizing sampling without replacement, and by using an appropriate normalization. In contrast to Poisson sampling, classical deviation inequalities in the i.i.d. setting do not straightforwardly apply to sample sums related to rejective schemes, due to the inherent dependence structure of the sampled points. We show here how to overcome this difficulty, by combining the formulation of rejective sampling as Poisson sampling conditioned upon the sample size with the Escher transformation. In particular, the Bennett/Bernstein type bounds established highlight the effect of the asymptotic variance of the (properly standardized) sample weighted sum and are shown to be much more accurate than those based on the negative association property shared by the terms involved in the summation. Beyond its interest in itself, such a result for rejective sampling is crucial, insofar as it can be extended to many other sampling schemes, namely those that can be accurately approximated by rejective plans in the sense of the total variation distance.