I tried to conduct a meta-analysis in both metafor from R and Review Manager from Cochrane using Standardized Mean Difference. While the SMD for each of the study is matching, the total effect size of the two software was a little different. I tried to apply Hedge G correction to metafor (As I checked the cochrane textbook it said that RevMan use Hedges correction for their total effect size) but the results still a little bit varied. Does anyone know the reasons? Thanks for the help
Specifically the effect size for the same outcome in Revman was
Hedge g = 1.25 [0.97, 1.54] Heterogeneity: Tau² = 0.55; Chi² = 1966.38, df = 27 (P < 0.00001); I² = 99%, Test for overall effect: Z = 8.63 (P < 0.00001)
while for metafor was:
tau^2 (estimated amount of total heterogeneity): 3.5079 (SE = 0.9658)
tau (square root of estimated tau^2 value): 1.8729
I^2 (total heterogeneity / total variability): 99.78%
H^2 (total variability / sampling variability): 459.47
Q(df = 27) = 1985.0169, p-val < .0001
Model Results:
estimate se zval pval ci.lb ci.ub
1.3222 0.3560 3.7143 0.0002 0.6245 2.0199 ***
