PCA score regression: the art of losing power

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lu, Yu, Pai, Nidhi, Cui, Erjia, Crainiceanu, Ciprian
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914594064171008
author Lu, Yu
Pai, Nidhi
Cui, Erjia
Crainiceanu, Ciprian
author_facet Lu, Yu
Pai, Nidhi
Cui, Erjia
Crainiceanu, Ciprian
contents The regression of principal component scores (RPCS) on covariates is a widely used analytic approach to detect and test for associations between functional measurements and study participant characteristics. Here we show that: (1) RPCS loses power relative to Function on Scalar Regression (FoSR); (2) the amount of power loss depends on the correlation between the PCs and the true effect; (3) if not corrected for multiplicity, RPCS has inflated $α$-level; and (4) current RPCS methods do not provide valid inference for the true effect. In contrast, we show that Function on Scalar Regression (FoSR) can avoid these problems using a particular combination of modeling tools. We validate these theoretical findings through extensive simulations and illustrate their practical implications using minute-level accelerometry data from the National Health and Nutrition Examination Survey (NHANES).
format Preprint
id arxiv_https___arxiv_org_abs_2605_24118
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle PCA score regression: the art of losing power
Lu, Yu
Pai, Nidhi
Cui, Erjia
Crainiceanu, Ciprian
Methodology
The regression of principal component scores (RPCS) on covariates is a widely used analytic approach to detect and test for associations between functional measurements and study participant characteristics. Here we show that: (1) RPCS loses power relative to Function on Scalar Regression (FoSR); (2) the amount of power loss depends on the correlation between the PCs and the true effect; (3) if not corrected for multiplicity, RPCS has inflated $α$-level; and (4) current RPCS methods do not provide valid inference for the true effect. In contrast, we show that Function on Scalar Regression (FoSR) can avoid these problems using a particular combination of modeling tools. We validate these theoretical findings through extensive simulations and illustrate their practical implications using minute-level accelerometry data from the National Health and Nutrition Examination Survey (NHANES).
title PCA score regression: the art of losing power
topic Methodology
url https://arxiv.org/abs/2605.24118