Regression with multiple variables in Gretl

Hello everyone,

since I'm new to the forum I have some doubts whether I should post my problem here, but here it goes...

I'm looking to make a regression with a dependent variable "ownership", which is dependent from multiple variables such as "amount of people that are married" and "amount of people that are between 20 and 29 years old". The problem is that all these former variables are percentages of the total population, while I also need to include variables such as "disposable income", which is not a percentage but rather a proper value, e.g. € 30.000. Do I need to transform certain variables to make the regression work?

Secondly, my promotor has hinted that I have to use panel data. Since I've only learned to make regressions with time-series data, I'm wondering if the same methods apply for panel data, e.g. testing for unit root and testing for residual autocorrelation?

Thanks in advance