I have scores of data sets ready for meta-analyses. I would like to run them as separate meta-analyses because otherwise I will be mixing apples with oranges. In the code below, I successfully run a single meta-analysis with a moderator. metacor(rho,N,Study,data=Leadership,comb.fixed=F,prediction=T,byvar=Leadership$Moderator) To test my idea, I used the variable Moderator to try to produce separate meta analyses with either by or tapply, to no avail. The closest I got to what I need is the code below. It does produce separate output for each level of Moderator, but it just replicate the results without separating the data. That is, I have k=13, and I just get the same k=13 meta-analysis for each level of Moderator. attach(Leadership) f<-metacor(rho,N,Study,comb.fixed=F,prediction=T) by(Leadership,Moderator,function(x) f) Yours, Avi Kluger<http://pluto.huji.ac.il/~mskluger/kluger.html> [[alternative HTML version deleted]]