Hello Contributors,

I have tried to understand what SDM is doing under the hood to generate the funnel plots and have come up short trying to find an explanation on the forums or citations. One thing we want to understand is why there are more dots than studies in the funnel plot? Is it using multiple imputations to generate a permutation-like test? 

Thanks for your help!

Daniel Sazhin

Dear Daniel, 

The blur points on the funnel plot are indeed imputations.

The funnel plot explains the effect sizes depending on the sample size, and it usually appears like a triangle (studies with small samples lead to larger effect sizes). Here, you should check if the dots are distributed around similar effect sizes, or at least all on one side of 0 (eg. There is significant bias if some studies of small samples reported negative effect sizes, while others were positive, and the studies of large samples reported null effect).

Best, 

Lydia