Hello, 

I'm including SPMs for an SDM meta-analysis. I have access to both thresholded and unthresholded SPMs. Is it better to use unthresholded original maps where possible, i.e. P<.0001? If unthresholded maps are unavailable, is it ok to use thresholded maps, e.g. FDR-corrected? Is it ok to have studies with both thresholded and unthresholded maps in the same meta-analysis? 

Best,
Molly