bootstrap diagnostic tests
† · 2025
Read bootstrap diagnostic tests by † (2025) 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-2509-01351.txt.
Description
Statistical conclusions often rest on a Gaussian limit, and when that limit fails, the answers can mislead. Giuseppe Cavaliere, Luca Fanelli, and Iliyan Georgiev show that the bootstrap, a resampling method, can itself detect such failures. Their diagnostic tests compare the bootstrap statistic's distribution with the Gaussian shape a valid setup would imply.
The tests carry three traits: they cause no pre-testing bias, they use the same critical values across applications, and they consistently spot departures from the Gaussian limit. The method is computationally simple, needing only repeated draws of the bootstrap statistic. The authors derive conditions that let data randomness and bootstrap randomness mix correctly. Several scenarios where the Gaussian approximation may break show the tests' practical reach for honest inference.
Catalog record
| CREATOR | † |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-01351