Meta-Analysis in Social Science Research

Abstract

Combining results of large number of heterogeneous studies into easily inferable aggregate measure is a challenge in social science, though of paramount importance. Policies cannot be based on results of individual studies as the inferences are limited by the scope of the study. Often the studies are based on smaller samples and restricted in terms of area coverage. Need of having a comprehensive aggregate indicator is even higher in cases where the studies give contradicting conclusions. The need for a method to synthesize results of numerous small studies into a more reliable, comprehensive statistical measure led to the development of ‘meta-analysis’. The meta-analysis approaches like vote counting method, combined probability method and modern meta analysis tools helps to summarize and combine the results of research on a particular area of study, to arrive at statistically valid overall effect size. It facilitates testing hypothesis over combined sample size which is more precise and account also for heterogeneity.Meta-analysis accompanied with systematic review not only emphasize its use in policy making, but also throw light on research gaps in a particular field of study.