The design with two sessions and two groups (treatment and control) is quite common. How can I calculate the confidence intervals for Cohen's d for that design?

It is said that in this design we should divide the mean differences by the pooled standard deviation (SDpooled) of the pretest groups (here refering to this). I was thinking to build the confidence intervals with this SDpooled, too. But I just read this:

The most appropriate standardizer is virtually always an estimate of the SD in the pretest population, perhaps $s_1$, the pretest SD in our data. By contrast, inference about the difference requires $s_{diff}$, the SD of the paired differences—whether for a paired t test or to calculate a CI on the difference. To the extent the pretest and posttest scores are correlated, $s_{diff}$ will be smaller than $s_1$, our experiment will be more sensitive, and a value of $d$ calculated erroneously using $s_{diff}$ as standardizer will be too large.

Does this mean that I calculate Cohen'd with the pooled SD of the pretest and the confidence intervals for this estimate with another pooled SD, namely with the pooled SD of the differences? This sounds odd to me since both results will differ as the cited text says, too.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.