Once you have decided to conduct a research synthesis, a second – and equally important – methodological choice awaits: should you perform a meta-analysis or a meta-synthesis? For many researchers, this distinction only becomes apparent mid-project, by which point a costly methodological mismatch may already have occurred. Understanding the difference from the outset will save you considerable time, effort, and potential revisions.
|
Why This Distinction Matters
The terms "meta-analysis" and "meta-synthesis" are both forms of synthesis conducted across multiple studies, which explains why they are so frequently confused. However, they serve fundamentally different purposes and are suited to entirely different types of data. Using the wrong approach is not merely a technical error – it can undermine the validity of your conclusions and invite serious scrutiny from supervisors, peer reviewers, and examiners.
|
What Is a Meta-Analysis? (Quantitative Pooling)
A meta-analysis is a statistical technique used to combine the numerical results of multiple quantitative studies addressing a similar research question. The process involves extracting comparable effect sizes – such as odds ratios, standardised mean differences, or correlation coefficients – from primary studies, then weighting and pooling them using fixed-effects or random-effects models.
|
The output is a single, statistically precise estimate that carries greater power than any individual study alone. Think of it as calculating a weighted average across studies: the more robust the study, the more it contributes to the overall result. Meta-analysis is inherently quantitative – if there are no numbers to pool, there is no meta-analysis.
|
What Is a Meta-Synthesis? (Qualitative Integration)
Meta-synthesis, by contrast, is an interpretive method used to integrate findings from multiple qualitative studies. Rather than combining numbers, the researcher identifies, analyses, and synthesises themes, concepts, and meanings across studies to construct a richer, more comprehensive understanding of a phenomenon.
|
Common approaches include meta-ethnography, thematic synthesis, and framework synthesis. The output is not a statistical estimate but a new, higher-order interpretation of qualitative evidence – one that goes beyond what any single qualitative study could offer. Meta-synthesis is used extensively in fields such as nursing, education, social work, and organisational behaviour.
|
Side-by-Side Comparison
|
Which One Answers Your Research Question?
|
The simplest decision rule is this: follow your data.
|
If your research question asks how much, to what extent, or what is the effect of, and your evidence base consists of comparable quantitative studies, meta-analysis is your method. If your research question asks what does it mean, how do people experience, or what are the processes underlying, and your studies are qualitative in nature, meta-synthesis is the appropriate choice.
|
A common point of confusion arises when a systematic review includes a mix of qualitative and quantitative studies. In such cases, neither meta-analysis nor meta-synthesis alone is sufficient. Researchers may instead opt for a mixed-methods synthesis, such as a convergent synthesis design, which integrates both forms of evidence while preserving their respective epistemological integrity.
|
Can You Combine Both in One Study?
Yes – and in some research contexts, doing so offers significant analytical depth. A researcher might conduct a meta-analysis of quantitative outcome data and a meta-synthesis of qualitative process data within the same broader systematic review. This approach is particularly valuable in healthcare and policy research, where understanding both whether an intervention works and why it works (or does not) is equally important.
|
Common Pitfalls
The most frequent mistake is attempting a meta-analysis when the available studies are too heterogeneous to pool – either statistically (high I² values) or conceptually (comparing fundamentally different interventions). In such cases, a narrative synthesis or meta-synthesis is more defensible. Equally, applying meta-synthesis to quantitative data is a methodological error that examiners and reviewers will identify quickly.
|
How GraceLitRev Supports Both Approaches
|
GraceLitRev is built to support researchers working across the full spectrum of synthesis methods. Whether you are extracting effect sizes for a meta-analysis or coding themes for a meta-synthesis, the platform's AI-assisted tools help you manage data extraction and organisation – ensuring your methodology is as rigorous as your research question demands.