How do I create a funnel plot for free?

Use the generator above: pick your effect measure (odds ratio, risk ratio, hazard ratio, mean difference, or SMD), add a row for each study with its effect size and standard error (or 95% CI), and the funnel plot renders instantly with the pooled-effect line, 95% pseudo-confidence funnel, and Egger’s test result. Download it as a scalable SVG — no sign-up required.

What is Egger’s regression test?

Egger’s test quantifies funnel plot asymmetry with a weighted linear regression: SND = a + b × precision, where SND = effect / SE (the standard normal deviate) and precision = 1 / SE. The intercept a measures asymmetry; it is tested against zero with a t test on n − 2 degrees of freedom. An intercept significantly different from zero (commonly p < 0.10) indicates funnel asymmetry, consistent with publication bias or other small-study effects.

How do I convert a 95% confidence interval to a standard error?

For difference measures (mean difference, SMD): SE = (upper − lower) / (2 × 1.96). For ratio measures (OR, RR, HR), take logs first: SE = (ln(upper) − ln(lower)) / (2 × 1.96), and analyze the effect as ln(ratio). Switch the input mode to "95% CI → SE" and the tool does this conversion for each study automatically.

Why is the standard error axis inverted?

The y-axis shows standard error with zero at the top, so large, precise studies (small SE) sit at the top of the plot and small, imprecise studies (large SE) sink to the bottom. With no bias, points spread out as you move down — forming the inverted-funnel shape that gives the plot its name.

How many studies do I need for a funnel plot?

As a rule of thumb, at least about 10 studies. With fewer studies the scatter is too sparse to judge symmetry visually, and Egger’s test has low power to detect real asymmetry. Cochrane guidance recommends against funnel-plot-based tests when a meta-analysis includes fewer than 10 studies.

Does funnel asymmetry always mean publication bias?

No. Asymmetry is consistent with publication bias but can also reflect true small-study effects — smaller trials may use different populations, more intensive interventions, or lower-quality methods that inflate effects — as well as selective outcome reporting or simply chance. That is why asymmetry tests are often called tests for small-study effects rather than tests for publication bias.

Research Tool

Funnel Plot Generator for Meta-Analysis

Q: How do I convert a 95% confidence interval to a standard error?

For difference measures (mean difference, SMD): SE = (upper − lower) / (2 × 1.96). For ratio measures (OR, RR, HR), take logs first: SE = (ln(upper) − ln(lower)) / (2 × 1.96), and analyze the effect as ln(ratio). Switch the input mode to "95% CI → SE" and the tool does this conversion for each study automatically.

Q: Why is the standard error axis inverted?

The y-axis shows standard error with zero at the top, so large, precise studies (small SE) sit at the top of the plot and small, imprecise studies (large SE) sink to the bottom. With no bias, points spread out as you move down — forming the inverted-funnel shape that gives the plot its name.

Q: How many studies do I need for a funnel plot?

As a rule of thumb, at least about 10 studies. With fewer studies the scatter is too sparse to judge symmetry visually, and Egger’s test has low power to detect real asymmetry. Cochrane guidance recommends against funnel-plot-based tests when a meta-analysis includes fewer than 10 studies.

Q: Does funnel asymmetry always mean publication bias?

No. Asymmetry is consistent with publication bias but can also reflect true small-study effects — smaller trials may use different populations, more intensive interventions, or lower-quality methods that inflate effects — as well as selective outcome reporting or simply chance. That is why asymmetry tests are often called tests for small-study effects rather than tests for publication bias.

Free funnel plot generator for meta-analysis — plot effect vs. standard error with a 95% pseudo-CI funnel and Egger’s test for publication bias. Export SVG.

Effect size vs. standard error — exact SVGRatio (log axis) & difference measuresPooled-effect line + 95% pseudo-CI funnelEgger’s regression test built in — free

Enter each study’s effect size and SE (or 95% CI) — renders an exact funnel plot as SVG, free

Effect measureTitle

Enter as:

Studies — effect & SE (log axis)

StudyEffectSE

Egger’s regression test

Intercept = -1.356 (SE 0.378), t(8) = -3.58, p = 0.007

Evidence of funnel asymmetry — possible publication bias or small-study effects.

Funnel Plot Examples

Exact engine renders plus AI illustrations of publication-bias checks

View:

Asymmetric Funnel (Egger Positive)

Exact engine render — ten trials with mild asymmetry; Egger’s test flags possible publication bias.

odds-ratioegger-testasymmetric

Symmetric Funnel (No Bias)

Exact engine render — a symmetric scatter consistent with no publication bias.

mean-differencesymmetricno-bias

Publication Bias Explained

AI illustration — how missing small studies create funnel asymmetry.

publication-biasexplainereducation

Risk Ratio on a Log Axis

AI illustration — a risk-ratio funnel plot on a log-scaled x-axis.

risk-ratiolog-axismeta-analysis

The Meta-Analysis Figure Set

AI illustration — the funnel plot alongside its companion PRISMA and forest plot figures.

systematic-reviewprismaforest-plot

SMD Funnel with an Outlier

AI illustration — an SMD funnel plot with one study outside the pseudo-confidence limits.

smdpsychologylinear-axis

What is a funnel plot?

A funnel plot is a scatter plot used in meta-analysis to detect publication bias and other small-study effects. Each study is one point: its effect size (an odds ratio, risk ratio, or mean difference) sits on the horizontal axis, and its standard error sits on the vertical axis — inverted, so large, precise studies appear at the top and small studies at the bottom. A vertical line marks the pooled effect, and two dashed lines form the 95% pseudo-confidence-interval funnel: the region where studies would be expected to fall if they differ only by chance. This generator draws that plot from your numbers and runs Egger’s regression test to quantify asymmetry.

How to read funnel plot asymmetry

With no bias, points scatter symmetrically inside the funnel — tightly at the top, widely at the bottom, like an inverted funnel.
A gap in one bottom corner — typically missing small studies with null or unfavorable results — suggests publication bias.
Asymmetry can also come from genuine small-study effects: lower methodological quality, different populations, or selective outcome reporting in small trials.
Check the Egger’s test result under the plot: an intercept significantly different from zero (p < 0.10) indicates statistically significant asymmetry.
A funnel plot needs enough studies to be meaningful — with fewer than about 10, the power to detect asymmetry is low.

How to make your funnel plot

Choose your effect measure — Odds Ratio, Risk Ratio, and Hazard Ratio use a log x-axis; Mean Difference and SMD use a linear x-axis.
Add a row for each study with its effect size and standard error — or switch to "95% CI → SE" and enter the confidence interval; the tool converts it with SE = (upper − lower) / (2 × 1.96), on the log scale for ratios.
The tool computes the fixed-effect inverse-variance pooled effect, draws the vertical line and 95% pseudo-confidence funnel, and runs Egger’s regression test automatically.
Set the title, then download a clean, scalable SVG for your paper, slides, or poster.

Publication bias and why it matters

Publication bias arises when the likelihood of a study being published depends on its results — trials with statistically significant, favorable findings are published faster and more often, while null or unfavorable small studies stay in file drawers. A meta-analysis that only sees the published subset overestimates the true effect. The funnel plot is the classic first-line check: if small studies cluster on the favorable side of the pooled effect with a corresponding gap on the other side, suspect missing studies. Egger’s regression test puts a p-value on that visual impression by regressing each study’s standardized effect (effect / SE) against its precision (1 / SE); a non-zero intercept signals asymmetry. For ratio measures, performing the test on the log scale — as this tool does — is the recommended approach.

Funnel plots alongside forest plots and PRISMA

A complete meta-analysis report pairs several figures. The PRISMA flow diagram documents how studies were identified, screened, and included. The forest plot presents each study’s effect size with its confidence interval and the pooled estimate. The funnel plot complements both by interrogating whether the included studies are an unbiased sample of all the evidence. Report the funnel plot whenever you have roughly ten or more studies, and state which asymmetry test you used. Build the matching figures with our forest plot generator and PRISMA flow diagram generator — all three export publication-ready SVGs.

Frequently Asked Questions

Related Research Tools

Research

Forest Plot Generator

Plot pooled effect sizes with confidence intervals for a meta-analysis.

Try it free

Research

PRISMA Flow Diagram Generator

Draw the PRISMA 2020 study-selection flow for a systematic review or meta-analysis.

Try it free

Research

CONSORT Diagram Generator

Build a CONSORT 2010 participant flow diagram for your randomized trial.

Try it free

View All Free Tools

Funnel Plot Generator for Meta-Analysis

Free funnel plot generator for meta-analysis — plot effect vs. standard error with a 95% pseudo-CI funnel and Egger’s test for publication bias. Export SVG.

Effect size vs. standard error — exact SVGRatio (log axis) & difference measuresPooled-effect line + 95% pseudo-CI funnelEgger’s regression test built in — free

What is a funnel plot?

How to read funnel plot asymmetry

With no bias, points scatter symmetrically inside the funnel — tightly at the top, widely at the bottom, like an inverted funnel.

A gap in one bottom corner — typically missing small studies with null or unfavorable results — suggests publication bias.

Asymmetry can also come from genuine small-study effects: lower methodological quality, different populations, or selective outcome reporting in small trials.

Check the Egger’s test result under the plot: an intercept significantly different from zero (p < 0.10) indicates statistically significant asymmetry.

A funnel plot needs enough studies to be meaningful — with fewer than about 10, the power to detect asymmetry is low.

How to make your funnel plot

Choose your effect measure — Odds Ratio, Risk Ratio, and Hazard Ratio use a log x-axis; Mean Difference and SMD use a linear x-axis.

Add a row for each study with its effect size and standard error — or switch to "95% CI → SE" and enter the confidence interval; the tool converts it with SE = (upper − lower) / (2 × 1.96), on the log scale for ratios.

The tool computes the fixed-effect inverse-variance pooled effect, draws the vertical line and 95% pseudo-confidence funnel, and runs Egger’s regression test automatically.

Set the title, then download a clean, scalable SVG for your paper, slides, or poster.

Publication bias and why it matters

Funnel plots alongside forest plots and PRISMA