ABSTRACT
This chapter describes meta-analysis and pointed out some of the limitations of these methods. It outlines United States efforts to conduct random assignment evaluations of mandatory welfare-to-work programmes and describes a unique database that was developed from information provided by these evaluations. Meta-analysis, widely used in medical and education research, provides a means for distilling the common key findings of an amalgam of policy evaluations. Meta-analysis is a quantitative approach, which consists of a family of statistical techniques that are applied to synthesize findings from multiple studies. Meta-analysis has two primary objects. The first is to provide statistically robust estimates of the size of a policy impact by pooling results across empirical studies, each of which may be based on a relatively small sample. Secondly, meta-analysis is used to explore characteristics that are associated with variations in the size of impacts observed in different settings and on differing occasions.
