Hi, I have a balanced panel data (cross-companies included 152 and periods 5 years). When I run panel unit root tests (LLC, Breitung, Im Pesaranm, ADF-Fisher, PP-Fisher, Hardi) I got different results for my variables (1 dependent, 8 independent variables). When I apply different tests, I always found that one independent variable is non-stationary. What is the best solution, which test could I apply (with lags or not, with the trend or not...). I use STATA.