由于在共变器上进行条件处理而导致的选择偏差。
Selection bias due to conditioning on a collider.
机构信息
CAUSALab and Departments of Epidemiology and Biostatistics, Harvard T H Chan School of Public Health, Boston, MA, USA.
Department of Communicable Diseases, National Centre of Epidemiology, Institute of Health Carlos III, 28029 Madrid, Spain
出版信息
BMJ. 2023 Jun 7;381:1135. doi: 10.1136/bmj.p1135.
Abstract
Effect estimates may be biased when the study design or the data analysis is conditional on a collider—a variable that is caused by two other variables. Causal directed acyclic graphs are a helpful tool to identify colliders that may introduce selection bias in observational research.
摘要
当研究设计或数据分析取决于一种由两个其他变量引起的“碰撞器”（collider）变量时，效果估计可能会产生偏差。有向无环图（Causal Directed Acyclic Graphs）是一种有用的工具，可以识别可能在观察性研究中引入选择偏差的碰撞器。
Zotero 插件