A beginner’s guide to causal inference methods: randomized controlled trials, difference-in-differences, synthetic control, and A/B testing
This text is meant for beginners who desire a comprehensive introduction to causality and causal inference methods, explained with minimal math.
On the subject of causality, we simply can’t avoid this classic statement: “Correlation doesn’t imply causation.” And a classic example is that simply because ice cream sales and drowning incidents are correlated, one doesn’t cause the opposite. You’ve probably heard many such examples illustrating the difference between the 2. While these examples are sometimes straightforward, the excellence can grow to be blurred in actual analyses.
With out a clear understanding of how causality is measured, it is simple to make incorrect causal inferences. On this regard, one query I often encounter is, “Yes, we all know that correlation doesn’t mean causation, but what a few regression evaluation?”. The short answer is that linear regression, by default, doesn’t provide any causal statements unless we undertake appropriate steps — that is where causal inference methods come into play.
