Hello Everyone,

I am working on a meta-analysis and would like to add a covariate (effect of smoothing kernel) to the linear model. In the outputs after FWE, I do not see outputs that represent the variance that is unique to the estimates associated with my variable of interest and the smoothing kernel covariate. 

Is there a way to separate out/threshold images that show only the main effect and only the covariate?

Thanks for your help!

Daniel

Dear Daniel, 

 

Are you introducing the covariate in the "mean analysis". If that's the case, you are just controlling for its effect in the analysis, if you want a better estimation of the covariate effect, you should perform a meta-regression.

 

Best,

Lydia