arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.22668math.STstat.MEstat.TH
收藏

柯西组合检验的固定水平校准

Fixed-level calibration of the Cauchy combination test

Graduate School of Arts and Sciences, The University of Tokyo(东京大学大学院人文社会系研究科);Komaba Institute for Science(驹场理工学研究院)

Hirofumi Ota

AI总结:

研究柯西组合检验在固定显著性水平下的渐近精确性,发现原始CCT在固定水平下不精确,提出边界层校准CCT(BL-CCT)通过修正参考分布而非统计量实现渐近精确,并在多种备择假设下保持功效。

摘要:

柯西组合检验(CCT)被广泛使用,因为它能产生闭式组合$p$值,并且在宽依赖结构下当名义水平$\alpha\downarrow0$时渐近有效。我们研究了一个不同的渐近问题:当组合$p$值的数量$K$在依赖下增长时,通常的柯西截断值在普通固定水平下是否仍然准确。在典型单因子等相关高斯copula模型下,我们证明原始CCT在固定$\alpha$下通常不是渐近精确的。在固定正相关下,统计量收敛到随机潜在因子极限,因此不存在通用的固定水平参考分布。当公共相关$\rho_K$随$K$减弱时,固定水平行为由边界层尺度$s_K=\sqrt{\rho_K}(\log K)^{3/2}$控制,且原始CCT渐近精确当且仅当$\rho_K(\log K)^3\to0$。由于大小失真完全来自参考分布而非统计量,因此可以在不修改检验统计量本身的情况下进行校正。我们提出了边界层校准CCT(BL-CCT),它用单参数高斯平滑柯西族替代标准柯西参考分布。与最近修改检验统计量的变体不同,BL-CCT保持统计量不变,仅校正参考分布。BL-CCT在更弱的条件$\rho_K\log K\to0$下渐近精确,并在有界边界层上提供有用的有限$K$近似。我们还进行了若干功效分析:尽管BL-CCT仅提高了截断值,但在局部密集、稀疏和密集高斯备择假设下,它在精确度尺度上相对于原始CCT没有一阶功效损失。数值实验支持校准理论。

英文摘要:

The Cauchy combination test (CCT) is widely used because it yields a closed-form combined $p$-value and is known to be asymptotically valid as the nominal level $α\downarrow0$ under broad dependence structures. We study a different asymptotic question: whether the usual Cauchy cutoff remains accurate at an ordinary fixed level when the number $K$ of combined $p$-values grows under dependence. Under a canonical one-factor equicorrelated Gaussian copula model, we show that the raw CCT is generally not asymptotically exact at fixed $α$. With fixed positive correlation, the statistic converges to a random latent-factor limit, so there is no universal fixed-level reference law. When the common correlation $ρ_K$ weakens with $K$, fixed-level behaviour is governed by the boundary-layer scale $s_K=\sqrt{ρ_K}(\log K)^{3/2}$, and the raw CCT is asymptotically exact if and only if $ρ_K(\log K)^3\to0$. Because the size distortion arises entirely from the reference law and not from the statistic, it can be corrected without modifying the test statistic itself. We propose the boundary-layer calibrated CCT (BL-CCT), which replaces the standard Cauchy reference by a one-parameter Gaussian-smoothed Cauchy family. Unlike recent variants that modify the test statistic, BL-CCT leaves the statistic unchanged and corrects only the reference law. BL-CCT is asymptotically exact under the weaker condition $ρ_K\log K\to0$ and provides a useful finite-$K$ approximation on bounded boundary layers. We also conduct several power analyses: although BL-CCT only raises the cutoff, it incurs no first-order power loss relative to the raw CCT on the exactness scale, under local dense, sparse, and dense Gaussian alternatives. Numerical experiments support the calibration theory.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑