Why it matters
Assignment by chance can help separate an intervention from pre-existing group differences. Good implementation and interpretation are still needed after randomisation.
A worked example
Eligible customers are randomly assigned to receive a reminder or not. Comparing average outcomes can estimate its effect under a valid design.
Illustrative example · simplified assumptionsA common mistake
Assuming random assignment makes findings representative of everyone.
Where the idea needs care
Non-compliance, attrition, spillovers and sampling limits still matter. Random assignment is not the same as random population sampling.
Apply the idea
Explain this concept using a different example from your spending, work, business or a policy debate. State what stays fixed and what could change the result.
Sources and further study
Examples and explanations by Alibomics. Numeric illustrations are not current market quotations.