Comparing control and treatment groups using regression and mixture models

doi:10.1080/10485252.2025.2591355

2025-12-01

PRE

Daniel R. Jeske *

DOI：10.1080/10485252.2025.2591355

原文链接

原文求助

摘要

En 中文

在比较治疗组与对照组的背景下，对治疗组的观测数据建立混合模型，可以推断治疗组中是否存在应答者和非应答者。模型中广义治疗效应由以下两个部分表示：一是患者为应答者的概率，二是控制组分布发生位移的幅度（该位移用于模拟对治疗有应答的患者反应）。文中推导了广义治疗效应的伪似然（PSL）和矩估计（MOM）估计量，这些估计量考虑了协变量的纳入。此外，还基于矩估计量（MOM）的渐近性质构建了置信区间。模拟结果显示，除非总体治疗效应较小，否则伪似然（PSL）估计量优于矩估计（MOM）估计量，且置信区间的名义覆盖率与实际覆盖率接近且令人满意。

Keyword:

Mixture model

covariates

responders

pseudolikelihood

nonparametric inference

期刊

Journal of Nonparametric Statistics

IF:

0.9

论文数:

被引数:

1.1K

机构

University of California System

学者数:

37.7W

论文数: 33.8W

被引数: 6.6K

引用论文

Statistical Inference for Method of Moments Estimators of a Semi-Supervised Two-Component Mixture Model

The American Statistician

IF0

2022-10-02

PREAI

Lubich,Bradley; Jeske,Daniel; Yao,Weixin

Novel active agents in patients with advanced NSCLC without driver mutations who have progressed after first-line chemotherapy 新型活性药物用于无驱动基因突变、经一线化疗后疾病进展的晚期非小细胞肺癌患者

ESMO Open

IF0

2016-01-01

OAAI

Christian Manegold; Alex Adjei; Federico Bussolino; Federico Cappuzzo; Lucio Crino; Rafal Dziadziuszko; David Ettinger; Dean Fennell; Keith Kerr; Thierry Le Chevalier; Natasha Leighl; Mauro Papotti; Luis Paz-Ares; Maurice Pérol; Solange Peters; Robert Pirker; Elisabeth Quoix; Martin Reck; Egbert Smit; Everett Vokes; Nico van Zandwijk; Caicun Zhou

Electronic Communications and Home Blood Pressure Monitoring (e-BP) study: Design, delivery, and evaluation framework

Contemporary Clinical Trials

IF0

2008-05-01

OAAI

Beverly B. Green; James D. Ralston; Paul A. Fishman; Sheryl L. Catz; Andrea Cook; Jim Carlson; Lynda Tyll; David Carrell; Robert S. Thompson

Logcondens: Computations Related to Univariate Log-Concave Density Estimation Logcondens: 与单变量对数凹密度估计有关的计算

JOURNAL OF STATISTICAL SOFTWARE

IF8.1

2011-01-01