中介分析
Mediation Analysis
| 節 | 週五 |
|---|---|
1 08:00–08:50 | 中介分析 YS107 2 節連堂 |
2 09:00–09:50 |
* 根據陽明交大上課時間表所列
本課程介紹反事實框架下的中介分析，涵蓋自然直接與間接效果、控制直接效果及介入式效果等概念，探討參數法、巢式參數法、加權、插補與自然效應模型，並延伸至多重中介。課程亦包含敏感度與差異分析，以檢驗假設穩健性並解析群體差異，並透過R語言實作結合理論與應用。 This course presents mediation analysis within the counterfactual framework, covering natural direct/indirect, controlled direct, and interventional effects. Methods include parametric, nested parametric, weighting, imputation, and natural effect models, extended to multiple mediators. Sensitivity and disparity analyses are introduced to assess robustness and group differences, with R applications linking theory to practice.
30% 出席與課堂參與，70% 報告/考試
| 週次 | 主題 |
|---|---|
| 第 1 週 | 1 Introduction and Background |
| 第 2 週 | 1 Counterfactual Framework |
| 第 3 週 | 2 Natural (in)direct Effects |
| 第 4 週 | 2 Controlled Direct Effect |
| 第 5 週 | 2 Interventional (in)direct |
| 第 6 週 | Case study, report/exam* |
| 第 7 週 | 3 Parametric / Nested Parametric Method |
| 第 8 週 | 3 Weighting / Imputation Method |
| 第 9 週 | 3 Natural Effect Model |
| 第 10 週 | Case study, report/exam |
| 第 11 週 | 4 Multi-mediator Problem |
| 第 12 週 | 4 Natural Effect Modeling |
| 第 13 週 | 4 Interventional Effect Modeling |
| 第 14 週 | 5 Sensitivity Analysis |
| 第 15 週 | 5 Disparity Analysis |
| 第 16 週 | Case study, report/exam |
1. Explanation in causal inference : methods for mediation and interaction, VanderWeele, Tyler J. 2015 2. Statistical Causal Mediation Analysis with R, Anning Hu, Springer, 2024
- 地點
- 醫學二館211
- 時間
- Email預約
- 聯絡方式
- ccwen@nycu.edu.tw