Well, the great Software Engineering As Science conundrum is that nobody can afford to run studies like that, let alone enough of them to get some sort of stati

Well, the great Software Engineering As Science conundrum is that nobody can afford to run studies like that, let alone enough of them to get some sort of statistically significant sample. So we just have to do our best.