Sunday, November 27, 2005

Something to read over a cigarette

Be sure to print off the entire article so you can read it as you shiver outside having your smoke:

Science suffered from a mass hypnosis. Studies over-estimated statistical co-relations. They diligently studied not what they claimed but something more accessible. Sugar-substitutes for humans were banned—based on white mice analogies. Theories on married couples arose—from studying college students. Systems of mass Darwinian behavior and forced altruism were propounded—from territorial behavior of monkeys and Professors. Many social science and medical studies had real confusions about basic statistical definitions. Did growing smoking studies have serious technical defects?

"You have no idea," he said." For there's no evidence that smoking causes anything in moderate doses but researcher's salaries. None. It may even extend life as part of a complex of healthy behaviors. It may be dangerous but life-style factors, e.g., poor diet, set it off. But no one wants to hear. Moderate smoking may harm, but evidence, statistically, is the reverse."
HT: Jomama of To herd or not to herd