Tools for Those Who Summarize the Evidence Base
I have recently been analysing the results of my meta-analysis of 102 interventions and just for fun I decided to run all the analyses using various model estimators to see what’s happening with the adjustment (are they more conservative/ less conservative etc.)?
In the random effects model, I’ve tested:
Restricted maximum likelihood (with and without the Knapp and Hartung adjustment)
I've attached the output and hope it may be helpful to someone, I find that the descriptors can be quite difficult to digest and a practical example might may it a little more clear.
*I don't use CMA, I use R, but tested the output from CMA on a complimentary trial