Calculating effect sizes after a diff-in-diff regression

[Aoife McNally]

Hi all

I was wondering if there's a way to calculate effect sizes (e.g. Cohen's d) after a diff-in-diff regression (specifically using the diff command) in Stata? TIA!

[Jared Greathouse]

You can calculate Cohen's D, but you're better off calculating partial correlations. Stata won't do it for you, but you can do it if you wanna write the code for it.

[Aoife McNally]

Thanks Jared, that's really helpful. If anyone already has code written out to calculate effect sizes post-DiD regression, I'd really appreciate it.

[Jared Greathouse]

There's almost certainly not a command, you'd need to get your hands dirty with Stata's matrix postestimation options. And to be fair, al you'd need is the t stat, observations in the model and number of predictors, so strictly speaking, it wouldn't be that big a deal, but there won't be a one line solution, is all I mean.