First Course Ding: Design And Adjustment (Chapters 5 To 8)

This section groups the design-adjustment chapters:

• Chapter 5: Stratification and Post-Stratification
• Chapter 6: Regression Adjustment and Rerandomization
• Chapter 7: Matched-Pairs Experiments
• Chapter 8: Unifying Fisherian and Neymanian Inference

The common theme is that design information and adjustment structure should enter the estimator explicitly rather than being treated as an afterthought.

1 What Is Covered

• blocked and post-stratified treatment-effect estimation
• Lin-style regression adjustment
• rerandomization as a restriction on the assignment mechanism
• matched-pairs randomization inference
• the difference between Fisher’s sharp-null test and Neyman’s average-effect logic