The two dangerous assumptions that are held in the research in gender different include the following. The assumption that, if we find a sex differences in some ability or kind of behavior, that means that all males do a particular thing and all females do some quite different thing ( for example, all men are aggressive, and all females are passive and peace-loving). Test scores of males and females overlap and thus, finding a sex difference does not mean that all women are one way and all men are one way. Therefore, knowing ones sex cannot possibly be used to predict the behavior of such an individual.
The second assumption is that sex differences are biologically based and, therefore, inevitable and unchangeable. Difference may arise not only based on the biology of the person but, also from the different ways girls and boys are raised and other differences that could have some biological basis like height that has been shown to be fairly easy to modify. The environment of genes affects them and this could bring differences too (Paula J. Caplan, Jeremy B. Caplan, 1994. P. 3).
Gender bias becomes a factor in research because most researchers base their researches on the two assumptions mentioned above. That leads to testing the hypotheses in a biased manner and consequently producing biased results. For example, if a Dr. John hypothesis is that men are stronger than women, he shall test this hypothesis based on jobs that require lots of energy by comparing the number of women and men in the jobs. This does not give the actual scenario that can lead to an unbiased report. Women who may be assumed to be less strong carry babies and vegetable and other heavy material for much longer time than men. Therefore if Dr. John just compares the workers in one institution, the result shall be biased.
References
Paula J. Caplan, Jeremy B. Caplan (1994). Thinking critically about research on sex and gender.
HarperCollins College Publishers.