crabbymetrics
  • Home
  • API
  • Binding Crash Course
  • Supervised Learning
    • OLS
    • Ridge
    • Fixed Effects OLS
    • ElasticNet
    • Synthetic Control
    • Logit
    • Multinomial Logit
    • Poisson
    • TwoSLS
    • GMM
    • FTRL
    • MEstimator Poisson
  • Semiparametrics
    • Balancing Weights
    • EPLM
    • Average Derivative
    • Double ML And AIPW
    • Richer Regression
  • Unsupervised Learning
    • PCA And Kernel Basis
  • Ablations
    • Variance Estimators
    • Semiparametric Estimator Comparisons
    • Bridging Finite And Superpopulation
  • Optimization
    • Optimizers
    • GMM With Optimizers
  • Ding: First Course
    • Overview And TOC
    • Ch 1 Correlation And Simpson
    • Ch 2 Potential Outcomes
    • Ch 3 CRE And Fisher RT
    • Ch 4 CRE And Neyman
    • Ch 9 Bridging Finite And Superpopulation
    • Ch 11 Propensity Score
    • Ch 12 Double Robust ATE
    • Ch 13 Double Robust ATT
    • Ch 21 Experimental IV
    • Ch 23 Econometric IV

On this page

  • 1 What Is Covered

First Course Ding: Observational Adjustment (Chapters 11 To 13)

This section groups the observational-study chapters:

  • Chapter 11: The Central Role of the Propensity Score
  • Chapter 12: Doubly Robust ATE Estimation
  • Chapter 13: ATT and Other Estimands

These are the chapters where the current crabbymetrics stack is most different from the original notebooks in a good way. The old third-party weighting and causal packages drop out in favor of:

  • Logit
  • BalancingWeights
  • AIPW
  • plain OLS or Ridge nuisance fits when needed

1 What Is Covered

  • overlap diagnostics and subclassification
  • entropy and quadratic balancing weights
  • manual and cross-fit AIPW logic
  • ATT versus ATE targeting