3 Applying causal inference

 

This chapter covers

  • Creating directed acyclic graphs to describe how data was generated
  • Make assumptions explicit with a directed acyclic graph
  • Graphing recommender systems, pricing, and marketing problems

In the previous chapter, we saw that we can model causal problems with graphs. We also learned that the graph is crucial for calculating the causal impact we are looking for. Depending on the geometry of the graph, we may need to apply the adjustment formula. Now we need to discuss how to create these graphs.

In this chapter, we explore some example scenarios where you can apply causal inference techniques. When tackling a specific problem with causal inference, there are two main steps. First you need to translate your problem description into causal language. Graphs are excellent tools for this—they help put into a model all the information you have about how data was generated. Creating the graph is crucial because it determines which formulas you’ll use later, like the adjustment formula. Many people find this stage challenging and don’t know where to start. That’s why we’ll use examples in this chapter to guide you through the graph creation process and offer tips to help you start your own graphs.

Once you’re comfortable with the graph that represents your problem, the second phase is to use formulas, algorithms, or other methods to estimate causal effects. The following chapters in this book focus on this phase.

3.1 When and why to use graphs in causal inference analysis

3.2 Steps to formulate your problem using graphs

3.2.1 List all the variables

3.2.2 Create your graph

3.2.3 State your assumptions

3.2.4 State your objectives

3.2.5 Check the positivity assumption

3.3 Other examples

3.3.1 Recommender systems

3.3.2 Pricing

3.3.3 Simulations

3.4 Further reading

3.5 Chapter quiz

Summary