Definition

Randomized controlled trial (RCT)

An experiment in which participants are randomly assigned to intervention or control, making the groups comparable and causal inference possible.

Random assignment is the trick that balances everything — measured and unmeasured — across groups on average, so outcome differences can be attributed to the intervention. Blinding (participants, providers, assessors) guards against expectation effects; the control arm defines the counterfactual.

RCTs top evidence hierarchies for treatment effects but are not always feasible or ethical, and their tightly selected samples can limit generalizability. CONSORT is the reporting standard; preregistration of outcomes is expected.