Government Social Research
Last Updated: 16/9/2008
Statistical approaches (often meta-analysis) are currently the most developed method for combining findings from different studies, particularly for answering impact questions. However, quantitative data can also be used to answer other types of question and impact questions can incorporate qualitative data.
Reviewing studies for impact can be seen as a two stage process. The first stage involves a qualitative description of the studies included in the REA in order to understand the data, plan the quantitative data synthesis and prevent errors in its interpretation. The second stage involves conducting the meta-analysis.
The objective of a descriptive REA is to collate and present the extracted data in a manner such that information about the characteristics (based on the inclusion criteria) and results of the studies included in the REA, are summarised in a meaningful way. The Centre for Reviews and Dissemination (PDF, The University of York website) suggests a tabular framework to undertake descriptive synthesis.
Meta-analysis is a technique for analysing statistical data of controlled experimental research. It can be used for:
Meta-analysis is a technically complex exercise that requires a statistician skilled in carrying them out. If these skills are not available in the REA Team you should consult an expert at the beginning of the project to determine whether it will be feasible to undertake a meta-analysis within the time available for the REA.
The Centre for Reviews and Dissemination (The University of York website) has guidance on undertaking meta analysis as well as a list of meta analysis software
Synthesis of findings - qualitative