Causal Learning
Moad HANI
March 2020
Contents
1 Overview 2
2 Concepts of causality 3
2.1 Definition of causality . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Causal effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 Confounders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.4 Counterfactuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3 Conditions for causality 3
4 Problems in causal reasoning 3
5 Comparisons 3
5.1 Causal and non-causal systems . . . . . . . . . . . . . . . . . . . 3
5.2 Causation vs correlation . . . . . . . . . . . . . . . . . . . . . . . 3
5.3 identification and causal inference . . . . . . . . . . . . . . . . . . 3
5.4 invalidity vs ambiguity . . . . . . . . . . . . . . . . . . . . . . . . 3
5.5 Causal and statistical designs . . . . . . . . . . . . . . . . . . . . 3
5.6 Direct vs indirect effects . . . . . . . . . . . . . . . . . . . . . . . 3
5.7 Causal effect and confounding . . . . . . . . . . . . . . . . . . . . 3
6 Causal models 3
7 Inference algorithms 3
8 Experimentation 3
8.1 Laboratory vs Field experiments . . . . . . . . . . . . . . . . . . 3
8.2 Experimental vs Non-experimental designs . . . . . . . . . . . . . 3
8.3 Limitations of experimentation . . . . . . . . . . . . . . . . . . . 3
9 Summary 3
10 Key terms and concepts 3
1
1 Overview
The notion of causality plays an essential role in the ordinary, scientific and
philosophical understanding of the world. Every day we ask ourselves questions
about the causes or effects of this or that phenomenon - why the coffee maker
does not work, what consequences this or that policy will have on employment,
etc. We ask ourselves the following questions. Most sciences have as their main
goal the search for causes. And a number of fundamental philosophical concepts
- knowledge, action, moral responsibility, etc. - are also important. - are difficult
or even impossible to understand without recourse to the notion of causality.
Figure 1: Causality different from Correlation :)
2
2 Concepts of causality
2.1 Definition of causality
2.2 Causal effect
2.3 Confounders
2.4 Counterfactuals
3 Conditions for causality
4 Problems in causal reasoning
5 Comparisons
5.1 Causal and non-causal systems
5.2 Causation vs correlation
5.3 identification and causal inference
5.4 invalidity vs ambiguity
5.5 Causal and statistical designs
5.6 Direct vs indirect effects
5.7 Causal effect and confounding
6 Causal models
7 Inference algorithms
8 Experimentation
8.1 Laboratory vs Field experiments
8.2 Experimental vs Non-experimental designs
8.3 Limitations of experimentation
9 Summary
10 Key terms and concepts
References