Accurate and robust inference

Abstract Classical statistical inference relies mostly on parametric models and on optimal procedures which are mostly justified by their asymptotic properties when the data generating process corresponds to the assumed model. However, models are only ideal approximations to reality and deviations from the assumed model distribution are present on real data and can invalidate standard errors, confidence intervals, and p-values based on standard classical techniques. Moreover, the distributions needed to construct these quantities cannot typically be computed exactly and first-order asymptotic theory is used to approximate them. This can lead to a lack of accuracy, especially in the tails of the distribution, which are the regions of interest for inference. The interplay between these two issues is investigated and it is shown how to construct statistical procedures which are simultaneously robust and accurate.

[1]  H. Rieder.  Robust asymptotic statistics , 1994 .

[2]  Fallaw Sowell.  The Empirical Saddlepoint Approximation for GMM Estimators , 2006 .

[3]  Shinichi Sakata,et al.  HIGH BREAKDOWN POINT CONDITIONAL DISPERSION ESTIMATION WITH APPLICATION TO S&P 500 DAILY RETURNS VOLATILITY , 1998 .

[4]  I. Vajda,et al.  Convex Statistical Distances , 2018, Statistical Inference for Engineers and Data Scientists.

[5]  John Robinson,et al.  Nonparametric tests for multi-parameter M-estimators , 2017, J. Multivar. Anal..

[6]  Fallaw Sowell,et al.  The empirical saddlepoint estimator , 2019, Electronic Journal of Statistics.

[7]  Saddlepoint approximations for multivariate M-estimates with applications to bootstrap accuracy , 2008 .

[8]  Elvezio Ronchetti,et al.  Saddlepoint approximations and tests based on multivariate M-estimates , 2003 .

[9]  Elvezio Ronchetti,et al.  Robust Indirect Inference , 2003 .

[10]  Christian Gourieroux,et al.  Simulation-based econometric methods , 1996 .

[11]  S. Van Aelst,et al.  Principal Components Analysis Based on Multivariate MM Estimators With Fast and Robust Bootstrap , 2006 .

[12]  P. J. Huber.  Robust confidence limits , 1968 .

[13]  Jonathan B. Hill,et al.  Supplemental Material for GEL Estimation for Heavy-Tailed GARCH Models with Robust Empirical Likelihood Inference , 2015 .

[14]  E. Ronchetti,et al.  Saddlepoint tests for quantile regression , 2016 .

[15]  R. Dahlhaus,et al.  A frequency domain bootstrap for ratio statistics in time series analysis , 1996 .

[16]  S. Sheather,et al.  Robust Estimation and Testing , 1990 .

[17]  Stephane Heritier,et al.  Robust Methods in Biostatistics , 2009 .

[18]  Elvezio Ronchetti,et al.  Robust and accurate inference for generalized linear models , 2009, J. Multivar. Anal..

[19]  R. Tibshirani,et al.  An introduction to the bootstrap , 1993 .

[20]  A. Gallant,et al.  Which Moments to Match? , 1995, Econometric Theory.

[21]  E. Ronchetti,et al.  Robust inference with GMM estimators , 2001 .

[22]  William J. J. Rey,et al.  Robust statistical methods , 1978 .

[23]  Elvezio Ronchetti,et al.  Composite likelihood inference by nonparametric saddlepoint tests , 2013, Comput. Stat. Data Anal..

[24]  V. Yohai,et al.  Min-Max Bias Robust Regression. , 1989 .

[25]  Anna Clara Monti,et al.  On the relationship between empirical likelihood and empirical saddlepoint approximation for multivariate M-estimators , 1993 .

[26]  F. Trojani,et al.  Higher-Order Infinitesimal Robustness , 2012 .

[27]  Joseph P. Romano,et al.  Nonparametric confidence limits by resampling methods and least favorable families , 1990 .

[28]  Davide La Vecchia.  Stable Asymptotics for M‐estimators , 2016 .

[29]  Anthony C. Davison,et al.  Applied Asymptotics: Case Studies in Small-Sample Statistics , 2007 .

[30]  S. Lahiri.  Resampling Methods for Dependent Data , 2003 .

[31]  Elvezio Ronchetti,et al.  Empirical Saddlepoint Approximations for Multivariate M-estimators , 1994 .

[32]  G. Alastair Young.  ROUTES TO HIGHER-ORDER ACCURACY IN PARAMETRIC INFERENCE , 2009 .

[33]  H. Daniels.  Saddlepoint Approximations in Statistics , 1954 .

[34]  B. Efron,et al.  The Jackknife: The Bootstrap and Other Resampling Plans. , 1983 .

[35]  V. Yohai,et al.  Robust Statistics: Theory and Methods , 2006 .

[36]  Paul Rilstone,et al.  EDGEWORTH AND SADDLEPOINT EXPANSIONS FOR NONLINEAR ESTIMATORS , 2013, Econometric Theory.

[37]  P. J. Huber,et al.  Minimax Tests and the Neyman-Pearson Lemma for Capacities , 1973 .

[38]  Saddlepoint Approximations for Marginal and Conditional Probabilities of Transformed Variables , 1994 .

[39]  P. J. Huber.  Robust Estimation of a Location Parameter , 1964 .

[40]  John E. Kolassa,et al.  Series Approximation Methods in Statistics , 1994 .

[41]  R. Butler.  SADDLEPOINT APPROXIMATIONS WITH APPLICATIONS. , 2007 .

[42]  Howard Wainer,et al.  Robust Regression & Outlier Detection , 1988 .

[43]  Claudia Kirch,et al.  TFT-bootstrap: Resampling time series in the frequency domain to obtain replicates in the time domain , 2011, 1211.4732.

[44]  Gauss M. Cordeiro,et al.  A modified score test statistic having chi-squared distribution to order n−1 , 1991 .

[45]  D. Silva,et al.  Improved Score Tests for Generalized Linear Models , 1993 .

[46]  Aida Toma,et al.  Robust tests based on dual divergence estimators and saddlepoint approximations , 2010, J. Multivar. Anal..

[47]  Veronika Czellar,et al.  Accurate and robust tests for indirect inference , 2010 .

[48]  E. Ronchetti,et al.  Robust statistics: a selective overview and new directions , 2015 .

[49]  R. V. Mises.  On the Asymptotic Distribution of Differentiable Statistical Functions , 1947 .

[50]  F. Peracchi,et al.  Robust M-Tests , 1991, Econometric Theory.

[51]  Stephane Heritier,et al.  Saddlepoint tests for accurate and robust inference on overdispersed count data , 2017, Comput. Stat. Data Anal..

[52]  Susanne M. Schennach,et al.  Accompanying document to "Point Estimation with Exponentially Tilted Empirical Likelihood" , 2005, math/0512181.

[53]  G. Imbens,et al.  Information Theoretic Approaches to Inference in Moment Condition Models , 1995 .

[54]  Likelihood Ratio Specification Tests , 1997 .

[55]  Michel Broniatowski,et al.  Divergences and Duality for Estimation and Test under Moment Condition Models , 2010, 1002.0730.

[56]  P. J. Huber.  A Robust Version of the Probability Ratio Test , 1965 .

[57]  R. Spady.  Saddlepoint approximations for regression models , 1991 .

[58]  Elvezio Ronchetti,et al.  Small Sample Asymptotics , 1990 .

[59]  Yanyuan Ma,et al.  Saddlepoint Test in Measurement Error Models , 2011 .

[60]  Yuichi Kitamura,et al.  An Information-Theoretic Alternative to Generalized Method of Moments Estimation , 1997 .

[61]  F. Peracchi,et al.  Bounded-influence estimators for the tobit model , 1990 .

[62]  D. Cox,et al.  Asymptotic techniques for use in statistics , 1989 .

[63]  Aida Toma,et al.  Dual divergence estimators and tests: Robustness results , 2009, J. Multivar. Anal..

[64]  F. Hampel.  The Influence Curve and Its Role in Robust Estimation , 1974 .

[65]  E. Ronchetti,et al.  Robust Bounded-Influence Tests in General Parametric Models , 1994 .

[66]  Benjamin Holcblat.  On the Empirical Saddlepoint Approximation with Application to Asset Pricing , 2015 .

[67]  Young Min Kim,et al.  A frequency domain bootstrap for Whittle estimation under long-range dependence , 2013, J. Multivar. Anal..

[68]  E. Ronchetti,et al.  Saddlepoint approximations for short and long memory time series: A frequency domain approach , 2019 .

[69]  Laura Ventura,et al.  Between stability and higher-order asymptotics , 2001, Stat. Comput..

[70]  J. Tukey.  Configural polysampling , 1987 .

[71]  Werner A. Stahel,et al.  Robust Statistics: The Approach Based on Influence Functions , 1987 .

[72]  Howard D Bondell,et al.  Efficient Robust Regression via Two-Stage Generalized Empirical Likelihood , 2013, Journal of the American Statistical Association.

[73]  Wolfgang Härdle,et al.  On Bootstrapping Kernel Spectral Estimates , 1992 .

[74]  R. Zamar,et al.  Bootstrapping robust estimates of regression , 2002 .