APTS: Causal Inference Module
About
Aims of Course
History of Causal Inference
Philosophy
1
Introduction
1.1
Statistical vs causal questions
1.2
Target Trials
1.3
Target Trial Emulation
2
Frameworks for Statistical Causality
2.1
The ‘do’ approach
2.2
Potential Outcomes
2.3
Structural equation models
2.4
Decision theoretic approach
2.5
Finest fully randomized causally interpretable structured tree graphs
2.6
A hierarchy of the frameworks
3
Causal Effects
3.1
Effects in different populations
3.2
Conditional effects
3.3
Multiple time points
3.4
Mediation effects
3.5
Principal stratum effects
3.6
Other contrasts
4
Conditional independence
4.1
Independence
4.2
Conditional Independence
Bibliographic Notes
5
Graphical Models
5.1
Definitions
5.2
Markov Properties
5.2.1
Factorization and the local Markov property
5.2.2
Global Markov property
5.3
Paths and d-separation
5.4
Ancestrality
6
Causal graphical models
7
Structural causal models
7.1
Definition
7.2
Cross-world independences
8
Single World Intervention Graphs
8.1
SWIGs
8.2
Model Definition
8.3
d-separation in SWIGs
9
Causal assumptions
9.1
No unobserved confounding
9.2
Positivity
9.3
Consistency
9.4
No interference
10
Outcome Regression
11
Propensity scores
11.1
Balancing scores
11.2
Horvitz-Thompson estimator
12
Doubly Robust Estimation
12.1
Estimating equations
12.2
Augmented inverse probability weighting
13
Covariate adjustment
13.1
Parent sets
13.2
Back-door adjustment
13.3
Examples
14
Efficient adjustment
14.1
Optimal adjustment set
14.2
Other scenarios
15
Forbidden projection
15.1
Latent projection
15.2
Adjustment sets
16
Causal biases
16.1
Confounding bias
16.2
Selection bias
16.3
Bias due to conditioning on a mediator
16.4
Time-related biases
16.4.1
Time-dependent confounding
16.4.2
Attrition bias
16.4.3
Immortal time bias
16.4.4
Prevalent user bias
17
Marginal structural models
17.1
Estimation and inference
18
Machine learning approaches
19
Post-double selection
19.1
Frisch-Waugh-Lovell Theorem
19.2
Sparsity
19.2.1
Simulation
20
Double Machine Learning
20.1
Conditions for Double ML
20.2
Eligibility for 401(k) Example
21
Causal Forests
21.1
Random Forests
21.2
Causal Forests
21.3
Asymptotics
21.4
R Implementation
21.5
Burn the forest
22
Bayesian Additive Regression Trees
22.1
BART on our data
23
Meta-Learners
24
Time-varying confounding
25
Mediation
26
Instrumental variables
27
Causal discovery
References
Published with bookdown
Causal Inference
Chapter 27
Causal discovery