Abstract:Empirical-likelihood-based inference for parameters defined by the general estimating equations of Qin & Lawless (1994) remains an active research topic. When the sample size is small and/or the dimension of the accompanying estimating equations is high, the resulting confidence regions often have a lower than nominal coverage probability. In addition, the empirical likelihood can be hindered by an empty set problem. The adjusted empirical likelihood (AEL) tackles both problems simultaneously. However, the AEL confidence region with high-order precision relies on accurate estimation of the required level of adjustment. This has proved difficult, particularly in over-identified cases. In this article, we show that the general AEL is Bartlett-correctable and propose a two-stage procedure for constructing accurate confidence regions. A naive AEL is first employed to address the empty set problem, and it is then Bartlett-corrected through a resampling procedure. The finite-sample performance of the proposed method is illustrated by simulations and an example. The Canadian Journal of Statistics 43: 42-59; 2015 (c) 2014 Statistical Society of CanadaResumeL'inference parametrique basee sur la vraisemblance empirique telle que definie par les equations d'estimation de Qin et Lawless (1994) demeure un sujet de recherche actif. Lorsque la taille d'echantillon est faible ou que la dimension des equations d'estimation est elevee, les regions de confiance obtenues presentent souvent un taux de couverture inferieur a leur valeur nominale. De plus, le probleme de l'ensemble vide peut causer des difficultes. La vraisemblance empirique ajustee resout ces deux problemes simultanement. Les regions de confiance decoulant de ces resultats d'ordre superieur necessitent toutefois une estimation precise du niveau d'ajustement requis, ce qui s'avere difficile, surtout dans les cas de surspecification. Dans cet article, les auteurs montrent que la correction de Bartlett peut s'appliquer a la vraisemblance empirique ajustee et proposent une procedure en deux etapes pour la construction de regions de confiance precises. Ils utilisent d'abord une version nai ve de la vraisemblance empirique ajustee pour regler le probleme de l'ensemble vide, puis ils appliquent la correction de Bartlett a l'aide d'une methode de reechantillonnage. Ils illustrent la perfomance de leur methode sur des echantillons finis par des simulations et un exemple. La revue canadienne de statistique 43: 42-59; 2015 (c) 2014 Societe statistique du Canada

A Resampling Method Based on Pivotal Estimating Functions

A Simple Resampling Method by Perturbing the Minimand

The resampling method via representative points

Small Area Quantile Estimation

On Nonsmooth Estimating Functions via Jackknife Empirical Likelihood

Subsampling for Partial Least-Squares Regression Via an Influence Function

Sampling Importance Resampling Algorithm with Nonignorable Missing Response Variable Based on Smoothed Quantile Regression

A Monte Carlo Method for Variance Estimation for Estimators Based on Induced Smoothing.

Nearly optimal capture-recapture sampling and empirical likelihood weighting estimation for M-estimation with big data

Nearly Optimal Two-step Poisson Sampling and Empirical Likelihood Weighting Estimation for M-estimation with Big Data

Resampling Calibrated Adjusted Empirical Likelihood

A Perturbation Method for Inference on Regularized Regression Estimates

Improving Semiparametric Estimation by Using Surrogate Data

A Unified Approach to Semiparametric Transformation Models under General Biased Sampling Schemes

Efficient Estimation of Nonlinear Finite Population Parameters Using Nonparametrics

Confidence Interval Estimation by a Joint Pivot Method

Estimation in hidden Markov models via efficient importance sampling

Simple resampling methods for censored regression quantiles

A robust and efficient estimation method for nonparametric models with jump points.

Sampling with replacement vs Poisson sampling: a comparative study in optimal subsampling

A Fast Imputation Algorithm in Quantile Regression