Hello everyone,

I try to analyze PISA 2018 data and, more specifically, I would like to conduct a stepwise regression with an interaction in the second model.
I am able to create a null model without the interaction by use of two discrete, independent variables: escs_cat (which contains several categories of the escs value) and spktestlang (Child speaks mostly the language of assessment at home as a dummy).
I use this command:
pisareg math escs_cat spktestlang

But when I include an interaction between the country and one of the two independent variables, I get an error:
pisareg math spktestlang escs_cat##ib1.cnt
Error: # invalid name

I tried to test the coding by use of one plausible value and it worked.
However, I know that this is not sufficient as a result.

Can somebody please help me?

Thank you very much in advance
Best,
Hannah