There's evidence that selection happens at the organismal level (duh) and at the species level, but nothing else. There's some weak evidence that it can happen at levels above species (genus, family, etc), but it's not super well supported.

The question of population-level selection (what you've called group selection) is more contentious, although it shouldn't be. The grandpa of our field, E.O. Wilson, whom we all adore and wish we could constantly hug, loves the idea. Sadly, evidence doesn't love it. Like, there's basically none. There's no real theoretical underpinnings that would make it possible, either, because there's just too much gene flow between demes (...partially isolated breeding populations) to allow selection to happen.

Indeed, it is.