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 : Politics for Pros- moderated -- Ignore unavailable to you. Want to Upgrade?

To: SmoothSail who wrote (342527)	1/11/2010 6:12:01 PM
From: ManyMoose	2 Recommendations Respond to of 793622

The problem with statisticians is that they do not speak English. They speak a language called Statspeak. Statistics has its place, but it is a game of probability. Anybody who doesn't get that should not play Blackjack. There is a particular array in statistics called "The Poisson Distribution." The following clip from Wikipedia is in English, but don't bother to read the rest because it is in Statspeak. The Poisson distribution can be applied to systems with a large number of possible events, each of which is rare. (A plot of the Poisson distribution might look a little like a hockey stick with the handle projecting out to the right and the blade holding to the vertical axis of the grid. I am not implying any relationship between this particular statistical concept and the Global Warming debate.) But I digress. I employed the Poisson to evaluate seedling density in areas that were regenerating after being harvested. My possibilities for density were 0, 1, 2, 3, 4, 5 and up to 2000 seedlings per sample. It was very probable that I would have 0-15 seedlings in a particular sample but unlikely though possible that I would have, say 300. Suppose I took one sample and got the number 200, a very unlikely but possible result. Suppose further that I had no other opportunities for acquiring more samples, as in inferring temperature trends from tree ring growth. If I made my conclusions on the basis of that one sample, or on a few samples that stretched the possible but unlikely concept, I would be DEAD WRONG. Not just a little dead wrong, but catastrophically dead wrong. All these statistical models rely on assumptions that have limited reliability. They are mainly useful for understanding how things work and are less useful for predicting the future.