Behavioral Data Analysis with R and Python

Chapter 3. Introduction to Causal Diagrams

In fact, with few exceptions, correlation does imply causation. If we observe a systematic relationship between two variables, and we have ruled out the likelihood that this is simply due to a random coincidence, then something must be causing this relationship. When the audience at a Malay shadow theatre sees a solid round shadow on the screen they know that some three-dimensional object has cast it, though they may not know if the object is a ball or a rice bowl in profile. A more accurate sound bite for introductory statistics would be that a simple correlation implies an unresolved causal structure.

Bill Shipley, Cause and Correlation in Biology (2016)

Causal diagrams (CDs) may well be one of the most powerful tools for analysis most people have never heard of. As such they are one of the three extremities (vertices) of the causal-behavioral framework (Figure 3-1). They provide a language to express and analyze cause-to-effect relationships, which works especially well when dealing with behavioral data analyses.

Figure 3-1. The causal-behavioral framework for data analysis

In the first section of this chapter, I’ll show how CDs fit into the framework from a conceptual perspective, that is, how they are connected to behaviors and data. In the second section, I’ll describe the three fundamental structures in CDs: chains, forks, and ...