randomised

Randomised is the British English spelling of randomized, meaning determined by chance. In research and experimental contexts, randomised methods assign participants or units to different groups by chance rather than by choice. The aim is to create comparable groups and to allow valid causal inferences about the effect of an intervention.

In randomized controlled trials, randomisation is central. A random sequence is generated using computer algorithms or random-number tables. Allocation concealment prevents those recruiting participants from knowing the upcoming assignment, reducing selection bias. Blinding may also be used to minimize performance and assessment bias.

Common randomisation schemes include simple randomisation, where each participant has an equal probability of assignment; block randomisation, which ensures balance across groups within blocks; stratified randomisation, which randomises within subgroups such as age or sex; and minimisation, an adaptive method that balances multiple factors across groups.

Randomisation is a design principle in statistics and clinical research and is distinct from random sampling, which concerns how participants are selected from a population. Randomised designs improve internal validity but require adequate sample sizes and careful handling of dropouts and protocol deviations, with analyses often following the intention-to-treat principle.

Applications extend beyond medicine to psychology, education, agriculture, and social sciences, wherever researchers seek to evaluate causal effects under controlled conditions.

In computing contexts, randomised may describe algorithms or data structures that incorporate randomness to avoid systematic biases, reflecting the broader idea of introducing chance into design.