Synonym: effect size, standardized mean difference
Parameters:
(1) number of subjects in group A
(2) mean value for group A
(3) standard deviation for group A
(4) number of subjects in group B (often the placebo control)
(5) mean value for group B
(6) standard deviation for group B
pooled group variance =
= (((number in A) - 1) * ((standard deviation for A)^2)) + (((number in B) - 1) * ((standard deviation for B)^2))
total degrees of freedom =
= ((number in A) - 1) + ((number in B) - 1)
common standard deviation =
= SQRT((pooled group variance) / (total degrees of freedom)
standardized increment =
= ((mean for group A) - (mean for group B)) / (common standard deviation)
Interpretation:
• A value of 0 indicates no difference between the 2 groups.
• If a higher score indicates a better outcome, then a positive SI indicates that group A is better than group B.
• If a lower score indicates a better outcome, then a negative SI indicates that group A is better than group B.
