RIGOROUS METHODS OF IMPACT ASSESSMENT
The randomized controlled experiment is presented in this chapter as the best research design for the purpose of obtaining estimates of the net effect of an intervention. A detailed exposition of the randomized controlled experiment is provided in order to understand the limitations of other approaches. Approximations to randomized controlled experiments are also presented along with the conditions that permit their appropriate use. /> experimental and control groups are discussed. Special attention is paid to statistical controls, a method that is especially appropriate and useful for the evaluation of ongoing programs. The chapter ends with a recommendation for using both randomization and statistical controls, a mixed strategy that has the combined advantages of both approaches.