I'm looking for a way to check the non so strong correlation between my explanatory variables of my model.

I'm using panel data fixed effect, with a panel effect and not a real time effect.

Just to be clear, due to predict() problem on R with fixed effect of the plm() function, I'm using lm() with factor(Country) removing the constant.

To check the variable correlation, could I compute a correlation on all my data (without taking into account the time or the observation effect) or should I distinguish among my observations (here it's the Country) ?

Thanks for your time and tricks.