Funnel Plot for RVE in Stata?

[Lindsey Cunningham]

Hi everyone,
I am using robust variance estimation (robumeta in STATA) for my meta analysis and I am trying to assess publication bias. Does anyone know how we can create funnel plots that accounts for the nested data structures within each study? I am not finding any information online on how to create funnel plots for RVE (in STATA or R).

[George Ford]

maybe helpful: https://www.statalist.org/forums/for...alence-studies

[Tiago Pereira]

A simple approach, which is often used (but not the best one), is to perform a "within-study meta-analysis," combining the dependent effect sizes from the same study, and then constructing standard funnel plots/contour-enhanced plots. In randomized trials, the standard error of the correlated estimates should be inflated by a factor of sqrt(N/2) before the "within-study meta-analysis," where N is the number of groups. I am not sure if this would apply to your case, since the dependency structure was not clarified.

These references can also be useful:
https://www.tandfonline.com/doi/full...X.2021.1946835
http://www.danielgucciardi.com.au/up...a-analysis.pdf