I hope I'm not reading what you wrote uncharitably, but, by rewarding people with positive results, we are inevitably encouraging people to either design their

I hope I'm not reading what you wrote uncharitably, but, by rewarding people with positive results, we are inevitably encouraging people to either design their experiments in ways that confirm their hypothesis, or at worse, to falsify their data. Thankfully, I've seen very little of the latter, but I've seen quite a bit of the former. Even unconsciously, people are afraid to design experiments that would give a black and white answer and falsify a cherish theory. People with lots of positive results are not necessarily better scientists. They are just as often less scrupulous and better salesmen - Peter Thiel noticed that as well.