Free Republic
Browse · Search
News/Activism
Topics · Post Article

To: MontaniSemperLiberi
It is true that I am treating each race as an independent event, and not part of some larger regional trend. So why would the Central Limit Theorem apply here if I'm not breaking down some larger event into uncorrelated smaller events, that is, that these races are all local to begin with?

-PJ

22 posted on 10/30/2010 12:14:40 PM PDT by Political Junkie Too ("Comprehensive" reform bills only end up as incomprehensible messes.)
[ Post Reply | Private Reply | To 19 | View Replies ]


To: Political Junkie Too

Okay so you understand why the covariance is important. Have some fun and read this article. It’s a sobering lesson,

Recipe for Disaster: The Formula That Killed Wall Street
http://www.wired.com/techbiz/it/magazine/17-03/wp_quant?currentPage=all

The problem with applying the central limit theorem is that even if the covariance is zero (in which case the CLT can still apply), you don’t have an estimate of the variation, only the mean. The sampling error that the pollsters report is an estimate of the uncertainty of the mean. It is not an estimate of the variation.

Note that there are TWO kinds of statements that the pollsters make. One is an estimate of how individual races will go. The second is the total number of seats that will turn over. Each estimate requires TWO different models with different, proprietary, assumptions. You are mixing apples and oranges.


24 posted on 10/30/2010 12:32:47 PM PDT by MontaniSemperLiberi (Moutaineers are Always Free)
[ Post Reply | Private Reply | To 22 | View Replies ]

Free Republic
Browse · Search
News/Activism
Topics · Post Article


FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson