Dear Statalist - Community,

I conduct an analysis using a difference-in-differences estimation strategy. For my descriptive section, I have to check that my sample disaggregated by status (active, inactive or control) is not too different by nature.

Therefore, I used the following command to compare the mean values of my indenpendent variables by DiD status.

Code:
balancetable (mean if control == 1) (mean if active == 1) (mean if inactive == 1) (diff active if inactive ==0) (diff inactive if active ==0) (diff active if control !=1) x1 x2 x3 x4  ///
Yet, I am not quite sure how to proceed with my categorical variables such as education level for the comparison of categories between the active, inactive and control sample. I read some previous post about using the Chi2 test but am still not sure whether this is appropriate.

Code:
My current try is:
= > equal to (diff active if inactive ==0) in the balance table above:

tab education active    if inactive ==0 , chi2
Thank you very much for any help in advance!

Warm regards,
Bianca