Balancing and Elimination of Nuisance Variables
Balancing and Elimination of Nuisance Variables
复制标题
DOI：
10.2202/1557-4679.1209
复制
发表时间：
2010-01-01
影响因子：
1.2
通讯作者：
Asgharian, Masoud
中科院分区：
文献类型：
--
作者：
Noorbaloochi, Siamak;Nelson, David;Asgharian, Masoud
解决因果分析中的协变量不平衡问题将被重新表述为消除有害变量问题。我们展示了在反事实平衡环境中，如何使用平均、调节和边缘化技术来减少由于可能存在大量不平衡基线混杂因素而导致的偏差。讨论了 X 充足量和 X 辅助量的概念，并作为示例，展示了切片逆回归和来自回归理论的相关方法（估计中心充足子空间的基础）如何为基于倾向的分析提供替代总结。提供了指数族和椭圆对称分布族的示例。
Addressing covariate imbalance in causal analysis will be reformulated as an elimination of the nuisance variables problem. We show, within a counterfactual balanced setting, how averaging, conditioning, and marginalization techniques can be used to reduce bias due to a possibly large number of imbalanced baseline confounders. The notions of X-sufficient and X-ancillary quantities are discussed and, as an example, we show how sliced inverse regression and related methods from regression theory that estimate a basis for a central sufficient subspace provide alternative summaries to propensity based analysis. Examples for exponential families and elliptically symmetric families of distributions are provided.