There are other ways in which the incentives are misaligned. Granting agencies that prefer to fund researchers who nearly always "succeed" in proving their hyp

There are other ways in which the incentives are misaligned. Granting agencies that prefer to fund researchers who nearly always "succeed" in proving their hypothesis will find that research proposals are low-risk and low-information. At in at least one sense, the ideal experiment is one with a 50/50 chance of failing. Disconfirmation is also important.