A Meta-analysis is a research approach that takes a collection of primary research projects and combines their data in order to analyze the results across the combined data set. Simply said, a meta-analysis paints a big picture by linking multiple studies. Meta-analysis provides a broad and inclusive view of a topic by incorporating various studies and data.
Simply said, a meta-analysis paints a big picture by linking multiple studies.
The purpose of a meta-analysis is to try and discover if there is a shared “truth” or research finding across multiple studies. If several studies contribute data that lead to a common finding, it makes that finding more robust.
A meta-analysis involves complex statistical techniques. It is not simply looking at a bunch of similar studies and comparing the findings. It actually involves pulling the data from the various studies, combining the data and analyzing the results of the combined data.
As you can imagine, conducting a meta-analysis is a complicated and time-consuming effort.
First, the researchers have to identify relevant studies. This could involve looking at hundreds of studies and determining which ones cover the targeted topic.
Second, the researchers need to contact the researchers who conducted the initial, or primary, studies and ask for the raw data. Sometimes the primary researchers are not easy to contact or will not provide the data, and then they must pull summary data from published works instead.
Third, the search for primary data needs to extend beyond published research and this creates the “file drawer problem”. Thousands of research studies are conducted, but never published in academic journals – often because the data did not prove the intended hypotesis. These data sets are “put in a file drawer” instead of being published. But sometimes, a well-researched null result can be as important as a positive finding. This data is an important part of the broader story and should be included in the meta-analytic data set. The researchers have to do a treasure hunt to find these file drawer studies and collect the data.
Fourth, the researchers need to combine the data sets into one new, and very large, data file. Imagine the challenges here of taking a bunch of variables and data that were all developed separately with different assumptions and models and try to combine them into one coherent whole. Hard to do, but powerful when accomplished.
Finally, the meta-analysis researchers use advanced statistical techniques to analyze the new data set and reach conclusions. The large data set brings more power to the analysis and helps uncover smaller relationships. These conclusions help combine many smaller studies and set a new foundation for the topic.
To see an example of meta-analytic study, see the discussion of Equal Opportunity in Work and Family Solutions.
