Human beings are all too easily influenced by a thin veneer of rationality, and the use of a modicum of scientific jargon. Nazism's beliefs in the superiority of a particular race, which utilized corruptions of evolutionary theory, are a prime example of this. Likewise, its twin evil of communism, with its unshakable belief in certain models of human development, was quite willing to eliminate those who opposed any aspect of that approach. Such ideologies, for all their scientific aspirations, have more in common with the most primitive of religious beliefs than with any scientific field.
Excessive certainty is contrary to the very nature of scientific endeavors. A communist roommate once tried to "educate" me as to why Marx's science of human nature explained the past, present, and future of the human race. Clearly, in the face of the collapse of most communist states (they did wither away, though not as Marx predicted) a truly scientific movement or individual would be forced to reevaluate core beliefs. The fact that so many altered their views so little in the face of overwhelming experimental evidence is but one more nail in the coffin of communism. Radicals of all stripes-- religious fanatics, conspiracy theorists, fascists, xenophobes, and even obsessed Freudians, among others--are similarly convinced of the truth of their beliefs despite copious evidence to challenge their ideas. If you disagree with them, you fall into one of two categories: either you are a member of the enemy faction, or you have been sufficiently brainwashed (or repressed) that you have not yet learned that their movement holds the idea that will bring about utopia. (I strongly suspect that their utopias would be closer to Orwell's vision than Sir Thomas More's.)
Nor is it only radical ideologues who drape their research in the language of scientific sophistication only to subvert the core ideas of science. It is ironic that people whose aversion to scientific methodology contributes to their decision to pursue certain areas of study should then attempt to seek validation for their work by intimating that it has strong scientific underpinnings. Mainstream experts in the humanities and social sciences are often confused about when mathematical language is and is not appropriate. Citing statistics, graphs, diagrams, or mathematical formulae adds a mystique of authority to any viewpoint, particularly when one is communicating with audiences whose mathematical facility is no greater than that of the presenter.
Examples of this phenomenon surround every election campaign. During the 1992 American presidential race, a Yale professor claimed to have determined a formula that would predict the winner in future presidential elections, as well as the margin of victory. Utilizing quantitative measures of economic, social, and political factors, he predicted that George Bush would win re-election in the fall of 1992. People with a command of quantitative information could see the basic fault with this line of thinking: the professor was taking a handful of data points--the victors and margins of victory in a few elections--and using a multitude of parameters to fit a curve to these points. If the reader has a graphing program, they might try selecting a few random points and fitting a high-degree (i.e. multi-parameter) polynomial curve to them. One can easily design a curve which passes through the points, zigging and zagging as it goes, and which provides absolutely no insights either into what the points have in common or where future points might be placed. The professor, needless to say, was wrong; had he been right (a significant possibility in selecting a winner from among two leading candidates) his predictions would have been taken very seriously in the 1996 election.