You have to weigh the ends against the means. "Repeating the experiments because of sloppiness" is pretty much the bottom of the barrel when it comes to ends, in my opinion.