Pub. date: 2007 | Online Pub. Date: September 25, 2007 | DOI: 10.4135/9781412952637 | Print ISBN: 9780761923879 | Online ISBN: 9781412952637| Publisher:Sage Publications, Inc.About this encyclopedia
Anthony Petrosino & David Weisburd
Researchers often use a randomized experiment (also known as a randomized field or controlled trial, or a true or classical experiment) to evaluate the effects of a legal policy, practice, or program. Because they consider randomized experiments the most rigorous approach to such evaluation, they often refer to true experiments as the “gold standard.” A randomized experiment is an evaluation design that uses randomization (also known as random allocation or random assignment) to achieve statistically balanced groups. In the typical experiment, the researchers test the impact of an intervention on outcome measures of interest by comparing an experimental group that receives an intervention with a control group that does not. Experiments usually include individuals as the unit of analysis, but it is now becoming more common to see studies in which large aggregate units such as schools or crime hot spots are randomly assigned (also known as cluster randomized trials ...