I have a dataset with missings and I was told by my supervisor to run multiple imputation. This is done. Now I need to calculate Cohen's d effect size for the mean differences between the experimental group and the control group.
For example, what is Cohen's d with confidence intervals for this data:
Imputation 1 group value A 2 A 5 A 3 B 4 B 2 B 1 Imputation 2 group value A 2 A 5 A 2 B 4 B 3 B 1
Imputation 1: Mean difference of is
1 and pooled sd is
1.53. Cohen's d is
0.65 and confidence intervall is
-1.38 to 2.69.
For imputation 2: mean difference is
0.33 and pooled sd is
1.63. Cohen's d is
0.20 and confidence intervalls
-1.76 to 2.17.
I am not sure how to use Rubins rule here to get an overall Cohen's d and an overall confidence intervall.