Asymptotic well-calibration of the posterior predictive p p -value under the modified Kolmogorov-Smirnov test

Yueming Shen · 2025

Read Asymptotic well-calibration of the posterior predictive p p -value under the modified Kolmogorov-Smirnov test by Yueming Shen (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-2504-14077.txt.

Description

Yueming Shen and Surya Tokdar tackle, in a 2025 statistics paper named by its head as a study of a well-calibrated Bayesian p-value, a subtle flaw in a popular model check. The check reuses the same data twice, so its p-value may misbehave even in huge samples. Past work fixed the issue only for well-behaved test statistics.

The authors widen the fix to statistics that need not settle into a bell curve. They show that Kolmogorov-Smirnov type statistics meet the needed conditions. Such statistics are handy because they react to many kinds of model failure. Their p-values therefore end up properly calibrated as samples grow. Model checkers gain a firmer footing for a tool they use every day.

Catalog record

CREATORYueming Shen
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-14077

Browse the full library · About the Tower of Babel