er, forgive my ignorance but there can be -multiple- null hypothesis? I thought that just means there's no real cause/effect relationship and it's just random patterns incorrectly looking significant?