I believe that there are very few scientists who deliberately falsify their work, cheat on their colleagues, or steal from their students. On the other hand, I am afraid that a great many scientists deceive themselves from time to time in their treatment of data, gloss over problems involving systematic errors, or understate the contributions of others. These are the 'honest mistakes' of science. The scientific equivalent of the 'little white lie' of social discourse. The scientific community has no way to protect itself from sloppy or deceptive literature except to learn whose work is suspect as unreliable.
(1994, p. 44) cited in: Leonard Brand (1997) Faith, reason, and earth history - Integrity in Science" (1985)