The ideal of science are, of course, unbiased scientists. In reality, scientists are humans, therefore biased. Usually they are biased in favour of their own pet theories. Usually that's not a problem, as long as other scientists have other biases.
The more problematic variant is a common bias. In the case of sexual abuse, it is quite obvious that many researchers have a strong moral bias against sex with children. Some researchers don't deny their bias. For example, Finkelhor describes sexual abuse as a moral problem and compares volitional sexual relations with "consensual slavery".
Another remarkable quote from Finkelhor 1986, p.178 suggests not only that for many researchers political goals have priority, but that some of them go too far in their exaggerations even in Finkelhor's opinion:
An interesting article about biased scientists is Okami P., Sociopolitical Biases in the Contemporary Scientific Literature on Sexual Behavior with Children and Adolescents.
The influence of the bias on the results of the research is
dangerous, but even not the main problem. It appears sufficient to
use biased terminology to describe the results of the research:
Rind, Bruce; Bauserman, Robert
Biased terminology effects and biased
information processing in research on adult-nonadult sexual
interactions: An empirical investigation
Journal of Sex Research, 30(3), 260-269 (1993)
What seems really dangerous is a bias of "child abuse experts" if
they participate as experts in courtrooms:
Horner TM, Guyer MJ, Kalter NM
The biases of child sexual abuse
experts: believing is seeing
Bull Am Acad Psychiatry Law 21(3):281-292 (1993)