Researchers want a certain outcome, so they rig the game: skimping on the lit review, choosing methodology that’s most likely to yield the results they want, barely mentioning alternative explanations for their data, ignoring inconsistent findings, downplaying limitations of their research, and then exaggerating its significance.
On second thought, all research is desire-driven. Because behavior is necessarily goal-driven and you don't have goals without wanting something to happen and wanting is desire and doing science is a behavior. But some desires are more conducive to scientific progress than others. Like the desire for reality not to make fools of us.
But then there is vision-fueled desire, a vision of the good, the true, and the beautiful, all wrapped into one. The great destroyer of truth and lives.
Yeah, this is all very meta. For examples, I refer you to: here, here, here, and here.
For more on bias in academia, I refer you to:
Duarte, J. L., Crawford, J. T., Stern, C., Haidt, J., Jussim, L., & Tetlock, P. E. (2015). Political diversity will improve social psychological science. Behavioral and Brain Sciences, 38, 1-13.DOI: http://dx.doi.org/10.1017/S0140525X14000430 [and try this link, with no paywall, or this link to the preprint version]