Publication Bias Analysis
Publication Bias Analysis (Egger Test, Funnel Plot) · Also known as: Small-Study Effects Test, Funnel Plot Asymmetry Test, Egger Regression Test, Yayın Yanlılığı Analizi
Publication bias analysis examines whether the set of studies included in a meta-analysis is a representative sample of all conducted research, or whether studies with non-significant or unfavorable results have been systematically suppressed. Matthias Egger and colleagues introduced the regression-based funnel plot asymmetry test in 1997, providing a formal statistical complement to the graphical funnel plot inspection long used in evidence synthesis.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Apply publication bias analysis when conducting a meta-analysis with at least 10 primary studies, as the test has very low power with fewer studies. It is most appropriate when there is reason to suspect selective reporting or when the meta-analysis synthesizes studies from diverse publication channels. The test assumes that study precision is unrelated to true effect size in the absence of bias. Limitations include inability to distinguish publication bias from genuine small-study effects due to heterogeneity. Alternatives include the trim-and-fill method and the p-curve analysis.
Strengths & limitations
- Provides a formal statistical test for asymmetry, complementing subjective visual inspection of the funnel plot
- Simple to compute and widely implemented in standard meta-analysis software
- Quantifies the degree of intercept deviation, giving a sense of the magnitude of potential bias
- Applicable to a broad range of effect size metrics including odds ratios, risk ratios, and standardized mean differences
- Requires a minimum of approximately 10 studies for adequate statistical power; results are unreliable with fewer studies
- Cannot distinguish between true publication bias and other sources of small-study effects such as clinical heterogeneity or methodological differences
- May produce false positives when effect sizes are genuinely correlated with study size (e.g., in dose-response relationships)
- The test is sensitive to the choice of effect size metric; different scales can yield different conclusions
Frequently asked
How many studies do I need before running the Egger test?
A commonly cited minimum is 10 studies. With fewer studies the test is severely underpowered, meaning that even substantial asymmetry may yield a non-significant p-value. Cochrane Handbook guidelines recommend interpreting the test cautiously below this threshold and relying more on visual funnel plot inspection in such cases.
Does a significant Egger test prove publication bias?
No. A significant test indicates funnel plot asymmetry, which can arise from publication bias, but also from genuine heterogeneity where smaller, typically less rigorous studies report larger effects. Additional evidence such as grey literature searches, trial registries, and contour-enhanced funnel plots is needed to attribute asymmetry specifically to selective reporting.
What significance threshold should I use for the Egger test?
Egger et al. (1997) and subsequent methodologists recommend using a one-sided p < 0.10 threshold rather than the conventional 0.05, because publication bias testing is directional and the test has limited power. Some guidelines also suggest reporting the exact p-value and confidence interval for the intercept to allow readers to judge the evidence themselves.
Sources
- Egger, M., Davey Smith, G., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629–634. DOI: 10.1136/bmj.315.7109.629 ↗
How to cite this page
ScholarGate. (2026, June 2). Publication Bias Analysis (Egger Test, Funnel Plot). ScholarGate. https://scholargate.app/en/meta-analysis/publication-bias-analysis
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Meta-RegressionMeta Analysis↔ compare