Skip to content
Prev 526 / 5632 Next

[R-meta] "impute-the-correlation + robustness + sensitive analysis" strategy

Dear Celia

One thing which you could do while you are waiting is to do the 
imputations assuming all the correlations are equal. If you do this for 
a range of plausible values it may be that the substantive result of 
your meta-analysis is unaffected and so you can be confident in your 
interpretation. Most techniques are much more sensitive to some inputs 
than to others.

Michael
On 23/01/2018 20:35, C?lia Sofia Moreira wrote: