🚀𝐀 𝐅𝐢𝐫𝐬𝐭 𝐂𝐨𝐮𝐫𝐬𝐞 𝐢𝐧 𝐂𝐚𝐮𝐬𝐚𝐥 𝐈𝐧𝐟𝐞𝐫𝐞𝐧𝐜𝐞 by Peng Ding🧠💡University of California Berkeley; 428-page PDF https://arxiv.org/abs/2305.18793

𝕿𝖆𝖇𝖑𝖊 𝖔𝖋 𝕮𝖔𝖓𝖙𝖊𝖓𝖙𝖘:

👉 Correlation, Association, and the Yule — Simpson Paradox

I. Introduction

👉 Correlation, Association, and the Yule — Simpson Paradox

👉 Potential Outcomes

II. Randomized experiments

👉 The Completely Randomized Experiment and the Fisher Randomization Test

👉. Neymanian Repeated Sampling Inference in Completely Randomized Experiments

👉. Stratification and Post-Stratification in Randomized Experiments

👉. Rerandomization and Regression Adjustment

👉. Matched-Pairs Experiment

👉. Unification of the Fisherian and Neymanian Inferences in Randomized Experiments

👉. Bridging Finite and Super Population Causal Inference

III. Observational studies

👉. Observational Studies, Selection Bias, and Nonparametric Identification of Causal Effects

👉. The Central Role of the Propensity Score in Observational Studies for Causal Effects

👉. The Doubly Robust or the Augmented Inverse Propensity Score Weighting Estimator for the Average Causal Effect

👉. The Average Causal Effect on the Treated Units and Other Estimands

👉. Using the Propensity Score in Regressions for Causal Effects

👉. Matching in Observational Studies

IV. Difficulties and challenges of observational studies

👉. Difficulties of Unconfoundedness in Observational Studies for Causal Effects

👉. E-Value: Evidence for Causation in Observational Studies with Unmeasured Confounding

👉. Sensitivity Analysis for the Average Causal Effect with Unmeasured Confounding

👉. Rosenbaum-Style p-Values for Matched Observational Studies with Unmeasured Confounding

👉. Overlap in Observational Studies: Difficulties and Opportunities

V. Instrumental variables

👉 An Experimental Perspective

👉. Disentangle Mixture Distributions and Instrumental Variable Inequalities

👉 An Econometric Perspective

👉. Application of the Instrumental Variable Method: Fuzzy Regression Discontinuity

👉. Application of the Instrumental Variable Method: Mendelian Randomization

VI. Causal Mechanisms with Post-Treatment Variables

👉. Principal Stratification

👉 Mediation Analysis: Natural Direct and Indirect Effects

👉. Controlled Direct Effects

👉. Time-Varying Treatment and Confounding

🚀《因果推理的第一门课程》丁鹏🧠💡加州大学伯克利分校；428页便携文档https://arxiv.org/abs/2305.18793

👉相关性、关联性和尤尔-辛普森悖论

一、简介

👉相关性、关联性和尤尔-辛普森悖论

👉 潜在结果

二、随机实验

👉 完全随机化实验和 Fisher 随机化检验

👉 完全随机实验中的奈曼重复抽样推断

👉 随机实验中的分层和后分层

👉 重新随机化和回归调整

👉配对实验

👉 随机实验中 Fisherian 和 Neymanian 推论的统一

👉 连接有限和超种群因果推理

三、观察性研究

👉 观察性研究、选择偏差和因果效应的非参数识别

👉 倾向得分在因果效应观察研究中的核心作用

👉 平均因果效应的双重稳健或增广逆向倾向得分加权估计量

👉 治疗单位和其他估计值的平均因果效应

👉 在因果效应的回归中使用倾向得分

👉观察研究中的匹配

四、观察性研究的困难与挑战

👉因果效应观察研究中的无混杂困难

👉 E值：具有未测量混杂的观察性研究中因果关系的证据

👉 具有未测量混杂的平均因果效应的敏感性分析

👉 用于具有未测量混杂的匹配观察研究的罗森鲍姆式 p 值

👉 观察研究的重叠：困难与机遇

五、工具变量

👉实验视角

👉解开混合分布和工具变量不等式

👉 计量经济学的角度

👉 工具变量法的应用：模糊回归不连续

👉 工具变量法的应用：孟德尔随机化

六、治疗后变量的因果机制

👉主要分层

👉中介分析：自然的直接和间接影响

👉受控的直接影响

👉 时变处理和混杂

Share & Translate: Chinou Gea (秦陇纪) @2023 DSS-SDS, IFS-AHSC. Data Simplicity Community Facebook Group. https://m.facebook.com/groups/290760182638656/ #DataSimp #DataScience #computing #program #AI #ArtificialIntelligence #causal #confounding #causalinference #causality