Jonas said that studies of homeopathy may have produced mixed results because of biases the researchers brought to their studies. Many investigators have likely set out to either prove or disprove the benefits of the practice, he noted, and may have discarded any work that did not fit their theory.
SOURCE: Annals of Internal Medicine 2003;138:393-399.
Shoorly not a mainstream scientist admitting that such a thing could happen. My favourite example of this is the Milliken lab book. For those who don't know Milliken performed a magnificent experiment at the start of the last century to measure the charge of the electron. His lab book of the experiment is full of results with notes alongside saying "something wrong with this one, disregard". The strangest thing is that he finished up with a pretty accurate result even after all the fudging of the results.