4. Causal Inference

Causal inference is the science of estimating impact. In this lesson, you'll learn about some of the most common types of evaluation designs to estimate impact and the tradeoffs of each.

When we talk about estimating impact, we are trying to understand what would have happened in the absence of the program, or the counterfactual. In this lesson, we'll discuss some non-experimental approaches to approximating the counterfactual.

Quiz 1

A randomized evaluation (also known as a randomized controlled trial or RCT) is the most rigorous way to estimate impact. In this lesson, we'll discuss how RCTs eliminate selection bias and variants of the RCT design.

Quiz 2

Randomization isn't always possible, in which case we may attempt a quasi-experimental design, which require us to make more assumptions. Two common quasi-experimental designs that we'll discuss in this lesson are difference-in-differences and matching.

Quiz 3

Many challenges can crop up in an evaluation that will threaten the internal validity of your design. In this lesson, we'll learn about three common threats: non-compliance, attrition, and spillovers, as well as mitigation strategies.

Quiz 4