ADJUSTED R squared RANDOM effects model

[Barbara Heerkens]

Hi all,

Thank you so much for your help so far!

I have a question regarding the R squared of an random effects model.
In other posts I already found out that for the R squared of a random model you take the 'R squared overall' measure.
However, how do I calculate or find the adjusted R squared?

Thank you!

[Carlo Lazzaro]

Barbara:
see https://www.statalist.org/forums/for...r2-in-xtreg-re.
However, in -.re- model you're usually more interested in the between R-sq.

[Barbara Heerkens]

Thanks for your reply.
When I click through I cannot find anything about the adjusted R squared.
So, am I correct when I interpret your answer as 'you rather look at the between R-sq instead of the adjusted R sq'?

[Carlo Lazzaro]

Barbara:
as per -xtreg- entry, it seems that - xtreg, re- does not calculate adj_R-sq- (see -xtreg- entry, page 436).
Anyway, you can look at this (hopefully more helpful) thread: https://www.researchgate.net/post/Do...model_in_STATA.
Your interpretation of my previous answer is correct.

[Barbara Heerkens]

Thank you Carlo. The thing is, I also use fixed effects in my regression and therefore I am not sure whether I can just use the between R squared?

Thank you for the link to the other thread. I see that I could calculate the Adjusted R squared using the formula Adj R2 = [1-(1-R2)*[(n-1)/(n-m-1)]]
However, do you know which R squared I should use to calculate it? my random effects formula looks like this:

xtreg leverage post##connected size_lag1 profitability_lag1 markettobook_lag1 cash_lag1 i.industry i.year, re(cluster id)

Thank you so much!

[Carlo Lazzaro]

Barbara:
you can get the adjusted-R2a for the overall R-sq only for all -xtreg- specifications.
Actually, -xtreg, re- does not stores this result, but you can get it by running -regress- with the same code (because overall-Rsq from -xtreg,re-=R-squared from -regress-):

xtreg leverage post##connected size_lag1 profitability_lag1 markettobook_lag1 cash_lag1 i.industry i.year, re vce(cluster id)
reg leverage post##connected size_lag1 profitability_lag1 markettobook_lag1 cash_lag1 i.industry i.year, vce(cluster id)
ereturn list
di  e(r2_a)

[Barbara Heerkens]

Thank you so much Carlo!!!!