I am doing a meta-analysis of studies comparing A and B. As some studies used different scales (measurement techniques), we have decided to pool all studies and used standardized mean difference (SMD). We would like to do a meta-regression, incorporating some covariates to analyze the difference. I am using "metafor" package in R to do that, but "metafor" does not allow to do a meta-regression with SMD? Is there any specific reason? I can do manually the calculation of SMD then analyze it as raw mean difference? Is that a valid approach?
Asked
Active
Viewed 11 times
1
-
`metafor` can do meta-regression with any effect size measure, including SMDs. – Wolfgang Feb 04 '22 at 12:21