Inference under First-Order Degeneracy
Xinyue Bei, Manu Navjeevan · 2026
Read Inference under First-Order Degeneracy by Xinyue Bei, Manu Navjeevan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-07377.txt.
Description
Standard inference breaks when a setting transformation has zero gradient, as in mediation analysis where an indirect effect is a product of coefficients. Xinyue Bei and one co-author study inference under first-order degeneracy, where plug-in estimators behave in ways that depend on nuisance values that cannot be estimated consistently.
They show the failure is intrinsic: near degeneracy, neither regular nor quantile-unbiased estimation is possible. Yet minimum-distance methods still deliver uniformly valid confidence intervals, with standard chi-square critical values under stated settings and a simple bootstrap otherwise. Stats experts get both a limit theorem and a workable method for a case where textbook inference silently fails.
Catalog record
| CREATOR | Xinyue Bei, Manu Navjeevan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 148 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07377