Or (4) they do believe the research, and they use it in their calculation that says that while they should err on the side of caution and should overdo force "a little bit" just in case, that they should be careful not go too far...
... which turns out to be the same calculation they used all along.