ABSTRACT
Meta-analysis is the systematic accumulation of findings from previous evaluation studies. The impact of gun control on homicides, the impact of welfare reform's transition to Temporary Assistance for Needy Families (TANF) on employment outcomes and on children, and the impact of citizen and stakeholder participation in regulatory program implementation on indicators of implementation success. Meta-analysis is different. It is far more systematic than a casual review of previous research. It is intended to meet the highest validity standards in its assessments of the studies that it reviews. If certain design flaws consistently overestimate program impact, then that knowledge informs the debate that often ensues when studies have conflicting findings. This information requires that six effect-size values for the same study are entered into the meta-analysis database. The regression might also find that as the number and quality of experimental and statistical controls increase, the effect size does, too, controlling for other variables.
