spec: add funnel-meta-analysis specification

[github-actions]

New Specification: funnel-meta-analysis

Related to #4590

---

specification.md

funnel-meta-analysis: Meta-Analysis Funnel Plot for Publication Bias

Description

A funnel plot used in meta-analysis to assess publication bias by plotting individual study effect sizes against their precision (typically standard error). Studies scatter around a summary effect line, with pseudo 95% confidence limits forming an inverted funnel shape. In the absence of bias, studies distribute symmetrically around the summary effect; asymmetry suggests publication bias or systematic heterogeneity. This is a standard tool in systematic reviews and Cochrane-style meta-analyses.

Applications

• Systematic review authors assessing whether small-study effects indicate publication bias across included trials
• Cochrane reviewers generating standard funnel plots as part of required reporting for intervention reviews
• Research methodologists evaluating selective reporting by visually inspecting funnel asymmetry
• Epidemiologists checking meta-analytic robustness before drawing pooled conclusions

Data

• effect_size (float) - Point estimate from each study (e.g., odds ratio, mean difference, risk ratio)
• std_error (float) - Standard error of each study's effect estimate
• study (str, optional) - Study label or identifier for annotation
• Size: 8-30 studies
• Example: Meta-analysis of 15 randomized controlled trials comparing drug vs placebo, with log odds ratios and standard errors

Notes

• Y-axis shows standard error (inverted so that larger/more precise studies appear at the top)
• X-axis shows the effect size measure
• Draw a vertical line at the summary/pooled effect size
• Draw pseudo 95% confidence limits as diagonal lines forming the funnel shape (summary effect +/- 1.96 * SE)
• Plot individual studies as points (optionally sized by weight or sample size)
• Include a vertical dashed reference line at the null effect (0 for differences, 1 for ratios) if different from summary effect
• Asymmetry in the scatter pattern suggests publication bias

---

Next: Add approved label to the issue to merge this PR.

---

🤖 spec-create workflow