DataTerminology.comNEW
Also known as: RCT
An experiment that randomly assigns participants to treatment or control groups to estimate causal effects while minimizing bias.
DESCRIPTION
A randomized controlled trial is a gold-standard experimental design for measuring causal effects by randomly assigning participants to intervention and control groups. Randomization balances known and unknown confounders, reducing selection bias. RCTs are widely used in medicine, public policy, online experimentation, and social sciences. Key practices include blinding, pre-registration, power analysis, and intention-to-treat analysis. Ethical oversight and informed consent are essential when human participants are involved. In digital contexts, RCTs underpin A/B tests and multivariate experiments. While highly reliable, RCTs can be costly, time-consuming, or infeasible in some settings, motivating quasi-experiments as alternatives.
ALTERNATIVE DEFINITIONS
Wikipedia
A randomized controlled trial (or randomized control trial; RCT) is a form of scientific experiment which aims to reduce bias when testing a new treatment.
Highlights bias reduction as the core purpose of randomization in trials.
Cochrane
A study in which people are allocated at random (by chance alone) to receive one of several clinical interventions.
Emphasizes allocation by chance and the clinical intervention context.
NCI (NIH)
A study in which the participants are assigned by chance to separate groups that compare different treatments.
Focuses on random assignment to compare treatments in clinical research.
NICE
A study in which people are randomly allocated to receive one of several clinical interventions.
Defines random allocation within a UK health-standards context.
RELATED TERMS
A/B Testing Causal Inference Statistical Power Random Sample Treatment Effect Placebo Intention-to-Treat Blinding Sample Size Hypothesis Testing
COMPETING TERMS
Observational Study Quasi-Experimental Design Natural Experiment
RELATIONS
- Narrower: A-B Test
KNOWLEDGE AREAS
Data Science Testing & Validation
TARGET AUDIENCE
Data Scientist AI Product Manager Data Analyst Data Strategy Consultant
EXTERNAL SOURCES
- Wikipedia - Randomized controlled trial
- Cochrane - Cochrane glossary definition
- NICE - Regulatory glossary entry
- Cochrane - Overview of evidence standards
- ClinicalTrials.gov - How clinical trials work
TOP QUESTIONS
What makes a Randomized Controlled Trial uniquely powerful for causal inference?
Randomization balances both known and unknown confounders, reducing selection bias. This design property strengthens internal validity and supports credible causal effect estimation.
Category: Concept
How is a Randomized Controlled Trial different from A/B testing?
A/B testing is an RCT applied to digital products, typically with shorter cycles and automated assignment. Both rely on randomization, but RCTs often involve clinical or policy interventions.
Category: Comparison
When is a Randomized Controlled Trial not appropriate?
RCTs may be unethical, impractical, or prohibitively costly for certain interventions. In such cases, quasi-experimental designs or observational studies are more suitable.
Category: Risks
Why is blinding important in a Randomized Controlled Trial?
Blinding reduces performance and detection biases by preventing participants and investigators from knowing treatment assignment, preserving trial integrity.
Category: Best Practices
How does statistical power affect a Randomized Controlled Trial?
Power determines the likelihood of detecting a true effect. Adequate sample size and expected effect size are essential to ensure interpretable results.
Category: Requirements
What is intention-to-treat analysis in a Randomized Controlled Trial?
Intention-to-treat analyzes participants in their assigned groups regardless of adherence, preserving randomization and providing conservative, policy-relevant estimates.
Category: Method
What are common mistakes when running a Randomized Controlled Trial?
Common mistakes include inadequate randomization, poor allocation concealment, insufficient blinding, underpowered designs, and outcome switching without pre-registration.
Category: Mistakes
How do ethics and consent apply to Randomized Controlled Trials?
RCTs require ethical review, informed consent, risk minimization, and monitoring for safety. Participant welfare and transparency are paramount.
Category: Ethics
How do I start planning a Randomized Controlled Trial?
Define the research question, outcomes, and effect size; conduct power analysis; register the trial; and implement robust randomization, allocation concealment, and blinding.
Category: Getting Started