If a question is simple, binary, unambiguous, clear-cut and easily - TopicsExpress

If a question is simple, binary, unambiguous, clear-cut and easily set, if it contains two or more clear views held by the majority and represents the common desires of all or most of the population of a given area, if enough people are asked...polls *can* and often do work. Some polls conducted by professional statisticians are subsequently hijacked and warped by the statistically unqualified mainstream media. These polls are in constant danger of being rejected or mocked by the population, yet are often the result of years of painstaking research and/or study. Any mistake or warping at The Sun cannot be laid at the door of Glasgow University, or at the door of any legitimate and reputable statistical institution. On the other hand, the neglect of a single but vital variable can prove fatal to an otherwise solid poll. The guy or gal who administered the poll will be kicking themselves but also wouldnt expect all polls or the science of polls to be roundly condemned because of their daft mistake. If you defend and see genuine practical purpose in polling, it means you probably understand the mechanics and track record of this type of statistical gathering. If you distrust and condemn all polls (polls mean jack-sh$t), you should (and must) be able to defend your position by drawing attention to the mechanics behind what you perceive to be complex and widespread statistical fraud. Considering that the Referendum was polling at 46% (Yes) and 54% (N0) shortly before a virtually identical official result, would anyone here be willing to debate the value and function of good polling? Can anyone here properly explain their condemnatory stance? FIGHT! :D