总结：Chapter 15 - Outcome Regression and Propensity Scores

英文总结
This chapter discusses outcome regression and propensity score methods, the most commonly used parametric approaches for causal inference. However, they are presented after IP weighting, standardization, and g-estimation (g-methods) because they are not generally applicable to time-varying treatments and have limitations in complex longitudinal settings. The chapter covers their implementation, assumptions, and pitfalls.