Quantile Mediation Analytics

Canyi Chen, Yinqiu He, Huixia J. Wang, Gongjun Xu, Peter X.-K. Song · 2024

Read Quantile Mediation Analytics by Canyi Chen, Yinqiu He, Huixia J. Wang, Gongjun Xu, Peter X.-K. Song (2024) 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-2412-15401.txt.

Description

Canyi Chen and colleagues built a mediation analysis for quantiles in a 2024 statistics paper. Mediation asks whether an exposure reaches an outcome through an intermediate variable, and scientists often care about effects at specific quantiles, say the worst-off tenth, rather than the average. Theory for quantile mediation lagged behind practice, leaving users without guarantees, so the authors define natural direct and indirect effects at each quantile inside the counterfactual framework, both with closed forms.

Testing for no mediation is tricky because that null hypothesis is composite, mixing several ways mediation can vanish. Their adaptive bootstrap controls false positives properly, backed by theory and numerics. Worked examples show the method answering subgroup questions means-based mediation cannot phrase.

Catalog record

CREATORCanyi Chen, Yinqiu He, Huixia J. Wang, Gongjun Xu, Peter X.-K. Song
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE64 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-15401

Browse the full library · About the Tower of Babel