Propensity Score Matching

Developed by Paul R. Rosenbaum and Donald B. Rubin (The Central Role of the Propensity Score in Observational Studies for Causal Effects, 1983).

Pairs each unit that received a treatment with a similar untreated one, matched on its estimated chance of being treated, so the groups compare as if assigned at random.

When to reach for it

Estimating what a program, policy, or feature actually did, with data only from people who chose it themselves.

How it runs

1. Estimate each unit's chance of treatment from its observed traits.
2. Match each treated unit to untreated units with similar scores.
3. Check that the matched groups balance on observed traits.
4. Estimate the effect on the matched sample.
5. Test how sensitive the result is to unmeasured factors.

Where it sits in SOLVE eX

• 6.4 Conducting Follow-Up and Taking Feedback
• 3.2 Finding the Past Facts

Go to the source

• The Central Role of the Propensity Score in Observational Studies for Causal Effects, Biometrika, 1983