This question already has an answer here:
Good Morning, I have a doubt. I am doing a panel data model to calculate the benchmarking of some companies in the sector. is there any test that I can apply to know if I use fixed or random effects? I'm using eviews.