Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : Sharks in the Septic Tank -- Ignore unavailable to you. Want to Upgrade?

To: Dayuhan who wrote (17094)	6/19/2001 2:50:14 AM
From: Solon	Read Replies (1) \| Respond to of 82486

Excellent post, Steven. The power of religion runs deeper than the legal authority to light a huge fire--with you in it--much deeper than that...

To: Dayuhan who wrote (17094)	6/19/2001 11:31:24 AM
From: The Philosopher	Respond to of 82486

Correlation does not automatically indicate causation Absolutely. A principle which is far too often ignored, either inadvertently or, perhaps sometimes, intentionally. The abuse of statistics is, IMO, one of the most disturbing trends of contemporary discourse. Thirty years ago I designed and taught a course in high school titled "Mathimagination" one component of which was the abuse of statistics. The book I used for that portion of the course was the book I earlier recommended to Solon: Darrell Huff's "How to Lie with Statistics." It's a book which every educated person, IMO, should read. I also worked for a while designing polls and interpreting polling data. Fascinating work. I knew the theoretical dangers inherent in polling, but before I got down and dirty into it I hadn't realized how easy it is to manipulate polling questions and how difficult it actually is to create a truly unbiased, objective polling question. Just as one example: when given a range of options, people tend not to pick the end points. Say, for example, you ask "What in your opinion represents a poverty level income for a family of four" and give the options $0-$10,000, $10,000 to $15,000, $15,000 to $20,000, etc., If you give the same question to an equivalent group but give the ranges as $0-5,000, $5,000-10,000, etc. the second group is likely to give more answers in the combined choices of $0-5,000 and $5-10,000 than the first group gave in the range 0-10,000. So even if both groups should have the same beliefs about what constitutes a poverty level income, the second group is more likely to say that a $10,000 income is poverty level than the first group. Similarly, though in theory both groups should have the same number of people choosing the $10-15,000 option, in fact the first group will have a higher response rate in this range because it's the lowest non-end option. So just breaking the bottom range into two groups instead of one will change the results of the survey. That's a simplistic example, of course, but it just illustrates one way in which, if you wish to, and even if you don't wish to but just don't understand polling, you can skew your results. Statistics are an enormously complex topic, and very few schools teach students anything about it at all. So they go into the marketplace of ideas completely unarmed in this regard. Which is a major problem, IMO, because as with anything else, bad statistics will drive out good, to the detriment of the public discourse.