ABSTRACT

Simple experimental studies randomize study participants into two groups: a treatment group that includes participants who receive the offer to participate in a program or intervention, and a control group that includes participants who do not receive that offer. Such studies primarily address questions about the program impacts on the average outcomes of participants. While the corresponding results are highly policy relevant as they can be used to make inferences about the effectiveness of the program or intervention as it is offered, the average impact estimate often masks the fact that it is actually averaged across program units or sites which often implement the intervention to varying degrees of fidelity.