NCJ Number
238597
Date Published
March 2012
Length
10 pages
Annotation
By providing examples from diverse program areas, this report illustrates the feasibility and value of low-cost randomized controlled trials (RCTs).
Abstract
A RCT is a study that randomly assigns individuals or other units to one group that is eligible to participate in a program, or to a "control group" that is not. The unique advantage of this process is that, with a sufficiently large sample, it ensures the two groups are highly similar in characteristics (e.g., demographics and motivation). Thus, any difference in outcomes between the groups can be attributed to program participation and not to other factors. Other, more common methods of evaluation cannot by themselves provide definitive evidence of a program's impact. Even though RCTs are accepted as being the most reliable evaluation methodology, many believe they are too expensive to be practical in cost-cutting times. RCT costs can be reduced, however, by measuring outcomes using administrative data already collected for other purposes. These data may include student test scores, criminal arrests, and health care expenditures. As the quality and breadth of administrative databases have increased over time, opportunities to do low-cost RCTs have increased. Another way to reduce RCT costs is to use individuals or other units available to participate in the study without special recruitment efforts. Although low-cost RCTs may not be possible for the examination of all issues of interest regarding a program's effects, it can be used to determine the basic question as to whether or not the program improves participants' lives. This paper summarizes five well-conducted, low-cost RCTs conducted in community settings. The study costs ranged from $50,000 to $300,000, with random assignment itself composing only a small percentage of the costs. 12 references