As a first lesson in management, you need to know what you know and what you don't know.
Self-selected methodology is easy, but it's also wrong. You cannot draw ANY conclusions from that, because you cannot characterize sample biases. Safety attitudes at POA may or may not reflect the general population. You can't know that without a randomized sample of the whole population. You can't even draw conclusions about POA, because your respondents may be preferentially safety-conscious (i.e., unsafe pilots won't bother answering your survey), for instance.
You are at a top-tier public university. Surely you can find someone on that campus who understands the first thing about statistics to help you out. Hint: try a statistical science like psychology or astronomy.