摘要
建立强可忽略处理分配条件下因果推断的结构回归模型,估计平均处理效应.在正态分布假 设下,总体参数的极大似然估计是渐近正态无偏估计.提出的方法可推广到具有测量误差的一般情 形.
A structural regression model for causal inference is established to estimate the average treatment effect under strongly ignorable treatment assignment. Under the normal distribution, the maximum likelihood estimator for the population parameter is proved to be unbiased and asymptotically normal. The proposed method can be applied to the general case with measurement errors.
出处
《华南师范大学学报(自然科学版)》
CAS
2000年第4期7-12,共6页
Journal of South China Normal University(Natural Science Edition)
基金
国家自然科学基金重点项目!(39930160)资助课题
关键词
因果推断
强可忽略处理分析
结构回归模型
causal inference
strongly ignorable treatment assignment
structural regression model
average treatment effect
maximum likelihood estimator
asymptotically normal.