Glossary

Randomized controlled trial (RCT)

An experiment that randomly assigns subjects to treatment and control groups to isolate a causal effect.

Also called: RCT, randomized experiment

A randomized controlled trial randomly assigns participants to a treatment group, which receives a program, policy, or intervention, or a control group, which does not, so that on average the only systematic difference between the two groups is the treatment itself.

Because assignment is random, the two groups should be statistically similar on every measured and unmeasured characteristic before treatment begins, so any difference in outcomes afterward can be attributed to the treatment rather than to pre-existing differences between people who chose to participate and those who didn't. This is what separates an RCT from an observational study or a quasi-experimental method like difference-in-differences, which relies on statistical assumptions to approximate that comparability rather than achieving it by design. In digital products the same underlying logic appears as A/B testing, typically run against a holdout group that receives no change.

RCTs are considered the strongest available evidence for causal inference and are the gold-standard design within rigorous program evaluation, used to test policies from job-training programs to public-health interventions. Pitfalls include results from a trial in one place, time, or population not generalizing elsewhere; ethical or practical constraints that make randomization impossible, since a legally entitled benefit cannot simply be withheld from a control group; and trials too small to have the statistical power needed to detect a real effect.

Last reviewed September 22, 2026

In the index now

Related terms

Program evaluation Systematic assessment of whether a public or nonprofit program achieves its intended outcomes, and why. Causal inference The set of methods used to determine whether one variable actually causes a change in another, not merely correlates with it. Difference-in-differences (DiD) A causal method that compares the before-and-after change in a treated group against the before-and-after change in an untreated group. A/B testing A controlled experiment that compares two variants to measure which one produces a better outcome. Holdout group A segment of users deliberately kept out of a treatment, campaign or feature, so its true incremental impact can be measured against them.

Related guides

How to evaluate a public program Logic models, the counterfactual problem, and how RCTs and quasi-experimental methods answer "did this actually work" honestly.