Stage 6: Analysis & Synthesis: Step 1: What Types of Data Synthesis to Consider
Methods of Data Synthesis
There are a number of ways to synthesize the results from your included studies. In general, however, your data synthesis will either be qualitative, quantitative, or a combination of both.
Qualitative Syntheses
Qualitative SynthesesAs defined by the Ohio State University Library:
"A qualitative synthesis is a narrative, textual approach to summarizing, analyzing, and assessing the body of evidence included in your review." Ohio State University Library
The purpose of using a qualitative synthesis is to:
Qualitative syntheses are used when:
The sources included in your review are NOT similar (e.g., different study designs)
The sources included in your review use research designs that are not sufficient to support numerical analysis (e.g., qualitative research)
If the study's populations are different (e.g., clinical populations) and it is inappropriate to combine their data
The sources included in your review are NOT similar (e.g., different study designs)
The sources included in your review use research designs that are not sufficient to support numerical analysis (e.g., qualitative research)
If the study's populations are different (e.g., clinical populations) and it is inappropriate to combine their data Quantitative Syntheses
Quantitative synthesis, also known as meta-analysis or pooling, use a statistical approach to integrate and summarize the results of included sources. By combining data from multiple studies, you may then determine the overall effect of the variable under study (e.g., treatment).
The purpose of using a quantitative synthesis is to: (Higgins et al., 2015)
Several characteristics must be met in order to use a quantitative approach:
- The participant populations and research designs must be similar between data that are pooled
- The quality of the studies being pooled must be consistent
- There must be an adequate number of studies and data
- A review team member with statistical expertise or a statistical consultant is required
Ensuring that your sources are similar enough to analyze together is extremely important, as statistical pooling of diverse studies can be very misleading.
Mixed Methods
Your included studies may contain evidence that will allow you to analyze both qualitative and quantitative data. If this is the case, you will follow the steps to conduct both types of data syntheses.
Additional Resources