I would say that with sufficiently large sample group you would get the usual Gaussian curve. There would be people like Archimedes shouting "Don't disturb my circles!" on one extreme, you would get all those purely ideologically motivated “gender studies", eugenics, etc. on another extreme, but the belly would consist of science intertwined with ideology to various degrees -- either as bias (conscious or not) or as an excuse / means to make a particular research possible. I would also say that the more ideologically or socio-economically motivated the research would register, the more useless, forgettable, detached from reality, etc. its outputs would turn out to be (obviously).