A technique used to select participants randomly to avoid bias and increase the likelihood of generalizing the findings.