Dear all, 

I am preparing an analysis plan for a mega-analysis (meta-analysis with de-identified individual participant data). I have been advised to include an a priori power analysis to calculate the required sample size per group to achieve 80% power using the G*Power power analysis tool.

My question is - can I use SDM estimates, reported in a previous meta-analysis, as inputs in this power analysis tool? Can they be used in place of the usual Cohen's d? Are SDM estimates equivalent to a Cohen'd effect size (i.e. small (d = 0.2), medium (d = 0.5), and large (d = 0.8)?


Many thanks for your help with this matter!

Kind regards,
Sally

Dear Sally, 

You can use the output Hedges'g from the SDM, which are corrected Cohen's d; and in this case, they are equivalent to the effect size

Kind regards, 
Lydia

Dear Lydia, 

Thank you very much for your swift response!!

Are you able to tell me where in the SDM outputs the Hedges g can be found?

Kind regards,
Sally

Dear Sally, 

In the SDM window, once you have performed the meta-analyisis, you click in the "Extract" buttom, and select one desirable peak (you should do this for every significant peak). This will create a report with the Hedges'g 

Kind regards, 
Lydia