Here is what I have conjectured.
Let $H$ be a normal subgroup of $O(n)$ (the orthogonal group of an n-dimensional vector space). If there $\exists h\in H,\det h=-1$, then $H=O(n)$.
The cases $n=1,2$ are extremely easy to prove - only some simple algebraic calculations are required. However, I find that this approach doesn't seem to work. I am trying to do an induction on $n$, but this yields no results.
I could not find a counterexample either.
Question: Is the above statement true or false?
It fails for $n=3$. Just take $H=\{\operatorname{Id},-\operatorname{Id}\}$.