Skip to content
Prev 1654 / 5636 Next

[R-meta] metafor::rma-function: Statistically significant interaction, but increased tau2 – and how to get the slope from the output

Thanks for the feedback. Ok, then the last model is overparameterized, but 
this model:

rma.mv(logOR, logOR.var, mods = ~ CF1_Women..IMP,
        random = ~ CF1_Women..IMP | id, struct="GEN", data=d)

(i.e., random intercepts and slopes for the different meta-analyses) 
should be fine. But I understand your preference for sticking to the 
protocol and using fixed effects for meta-analyses and the interactions is 
also fine.

Best,
Wolfgang
On Fri, 2 Aug 2019, Sabrina Mai Nielsen wrote: