How to distort and bias polls and "studies"//If done in the open, imagine what goes on unseen.

Sotomayor's Focus on Race Issues May Be Hurdle by David Kirkpatrick, NY TIMES 5/30/09 ****In the course of this discussion, the NYTimes had occasion to quote polls and make their own inferences or endorse questionable ones by a pollster. "Her nomination and..have brought racial quotas back as a national issue,"...The article then says
"The public response, however, is hard to foresee. Few groups conducted public polls on the issue as it faded in recent years, and the results from those that did reval a consistent ambivalence, said Michael Dimock, a pollster with the nonpartisan Pew Research Center. When asked a question about "affirmative action or preferential treatment for minorities," the public has consistently opposed the idea by a margin of two to one. But when asked about "affirmative action programs designed to help women and minorities," an even bigger majority has supported them."...****But, but ...hold on. The first queried about minorities while the second cheerfully and carelessly added "women". Where's the common sense of equating these two questions and claiming contradictory answers? It seems obvious that non-minority women were likely less enthusiastic about advantaging minorities NOT including women than they were about advantages granted to WOMEN AND MINORITIES. This is not the logical "AND" (i.e. someone BOTH a woman and a minority ) but the inclusive one and the difference could well have been due to women ( and/or men ) thinking advantages to non-women minorities, or minorities in general, were a bad thing but were more generous about according advantages to women. The putative ambiguity seems in the mind of the polster trying to make an agenda point rather than objectively accumulating and interpreting data. In fact, how questions are phrased can probably change the results by as much as 180 degrees. Here's another "study" that reaches a politically correct conclusion, and might even be right, although probably on the basis of no statistically significant evidence **** http://tinyurl.com/kj4rdp
Girls worse at math? No way, new analysis shows By Maggie Fox, Health and Science Editor (Reuters) – Girls can do just as well at math as boys -- even at the genius level -- if they are given the same opportunities and encouragement, researchers reported on Monday. Their study, published in the Proceedings of the National Academy of Sciences, contradicts studies showing girls can do as well as boys on average in math -- but cannot excel in the way males can. They also said it is a clear rebuttal to Larry Summers, who as president of Harvard University said in 2005 that biological differences could explain why fewer women became professors of mathematics. ****Yoda wishes to say at the outset that he has no opinion on the matter, has three granddaughters and a daughter-in-law who has a PhD in mathematics from M.I.T. However, he believes Larry Summers was improperly pilloried for merely suggesting that the matter be studied since advances in neuroscience have indicated both structural and functional complexities and gender differences not known in the past.**** "We conclude that gender inequality, not lack of innate ability or 'intrinsic aptitude', is the primary reason fewer females than males are identified as excelling in mathematics performance in most countries, including the United States," Janet Hyde and Janet Mertz of the University of Wisconsin in Madison wrote in their report.... "We asked questions about how well females relative to males are doing at the average level, at the high-end level -- 95th percentile or above -- and the profoundly gifted level, the one-in-a-million type level,"...****The issue devolves on the level of "one-in-a-million". It is really doubtful that any statistically significant data existed about these rare occurrences in the "studies" done which seem to have asked "questions" of people rather than conducting any quantitative studies of ability at this extreme level. Since ideas about "rare genius" are starting to be questioned ( and actually to result from social stimulus and practice rather than only raw talent ), it is hard to conclude anything at this stage. One cannot imagine, however, any funding to be available for a study debunking this one's conclusions: writing a proposal to test the hypothesis that males are superior in the more rarefied achievements in mathematics would not find a hospitable reception. Additionaly, Yoda always finds amusing, amidst such proofs of equality, statements of countervailing INequality like "Analysis of data...indicated that as many, if not more girls than boys scored above the 99th percentile ...****