English

Double-estimation-friendly inference for high-dimensional misspecified models

Statistics Theory 2022-05-20 v3 Statistics Theory

Abstract

All models may be wrong -- but that is not necessarily a problem for inference. Consider the standard tt-test for the significance of a variable XX for predicting response YY whilst controlling for pp other covariates ZZ in a random design linear model. This yields correct asymptotic type~I error control for the null hypothesis that XX is conditionally independent of YY given ZZ under an \emph{arbitrary} regression model of YY on (X,Z)(X, Z), provided that a linear regression model for XX on ZZ holds. An analogous robustness to misspecification, which we term the "double-estimation-friendly" (DEF) property, also holds for Wald tests in generalised linear models, with some small modifications. In this expository paper we explore this phenomenon, and propose methodology for high-dimensional regression settings that respects the DEF property. We advocate specifying (sparse) generalised linear regression models for both YY and the covariate of interest XX; our framework gives valid inference for the conditional independence null if either of these hold. In the special case where both specifications are linear, our proposal amounts to a small modification of the popular debiased Lasso test. We also investigate constructing confidence intervals for the regression coefficient of XX via inverting our tests; these have coverage guarantees even in partially linear models where the contribution of ZZ to YY can be arbitrary. Numerical experiments demonstrate the effectiveness of the methodology.

Keywords

Cite

@article{arxiv.1909.10828,
  title  = {Double-estimation-friendly inference for high-dimensional misspecified models},
  author = {Rajen D. Shah and Peter Bühlmann},
  journal= {arXiv preprint arXiv:1909.10828},
  year   = {2022}
}

Comments

To appear in Statistical Science

R2 v1 2026-06-23T11:24:08.025Z