Towertalk
[Top] [All Lists]

Re: [TowerTalk] Divining rods

To: towertalk <towertalk@contesting.com>
Subject: Re: [TowerTalk] Divining rods
From: David Gilbert <xdavid@cis-broadband.com>
Date: Wed, 01 Sep 2010 11:01:23 -0700
List-post: <towertalk@contesting.com">mailto:towertalk@contesting.com>
That's a carelessly considered statement.  Although there are of course 
bad practitioners who simply make mistakes, statistics itself is an 
extremely rigorous discipline.  Wrong answers from statistics almost 
always come from one of two failures:

1.  Not properly understanding the physical process to which statistical 
analysis was applied.  It happens all the time when people run a test 
without taking into account all the variables.  They test the variables 
they know about and apply statistics to the resulting data, but get 
wrong results because they didn't include all the relevant variables in 
their test.

2.  Having a pre-existing bias that skews the data sample or the test 
itself.  Most people are not truly unbiased about any topic, and many, 
many experiments are crafted (knowingly or unknowingly) to prove a 
result, not investigate it.

Dave   AB7E



On 9/1/2010 6:47 AM, Frosty wrote:
> Statistics is also calculated guess work. You can make statistics give any
> answer you want.
>
>
> Charles F. Frost
> Frosty K5LBU
> frosty1@pdq.net
_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>