In my last blog post, I reviewed what a synthesis was, including the definition, its main purposes, and its function within higher education. If you have not read that post yet, I highly recommend you do so. Within this post, I will center my focus on synthesis by writing about a meta-synthesis. The book Conducting Research Literature Reviews: From the Internet to Paper provides a definition for this type of synthesis. A meta-synthesis is described as “a rigorous method to interpret and combine results from qualitative studies.”
The book recommends a series of seven steps to create a comprehensive and organized analysis. Below are the steps in order as presented in the text:
- Clarify the objectives of the analysis- meta-analysis is a research method, so you need to know the objectives. This is so you can evaluate the appropriateness of both the included and excluded literature.
- Set explicit criteria for including and excluding studies- within this step, one is recommended to conduct a sensitivity analysis. This is a technique where a separate analysis of different quality studies is conducted.
- Describe in detail the methods used to search the literature- It is important to describe all of the databases and search terms used to obtain the literature, along with the inclusion of all potentially relevant studies. This is crucial to avoid publication bias, a term that is described as “a review that unfairly favors the results of published studies.” In order to avoid this bias, you need to find negative or in progress results.
- Search the literature using a standardized protocol for including and excluding studies- to ensure a consistent review, a screening protocol should be created.
- Use a standardized protocol to collect data from each study regarding study purposes, methods, and effects- valid data collection often requires at least two reviewers using a standard protocol.
- Describe in detail the statistical method for pooling results- one assumption when creating a meta-synthesis is that you can pool individual study results to produce a summary because all results reflect the same effect. If the assumption is correct, then when the results are combined any random errors will be canceled out, and one meta-study will be produced.
- Report results, conclusions, and limitations.