GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2004-01 > 1073031229


From: "Debby Peare" <>
Subject: [DNA] By George, I think I've got it! LOL
Date: Fri, 2 Jan 2004 03:13:49 -0500 (Eastern Standard Time)
References: <20040102063918.69841.qmail@web41214.mail.yahoo.com>


GOOD GRIEF! That first one was a mess!
Let's try this email ONE more time! LOL
**********************************************

Ok David F. Ray & Charles!
The light bulb has been lit and I
understand this now! As I've been
lying in my bed, I've done the math
work and see how the trianglation
and bar graph works. I understand
not only the percentage, but contours
and probabilities as well. And I've
got to say.... you 3 are really a bunch
of pessimists! LOL But, that's the
territory that usually comes with being
a scientist. *VBS*
Taking the LEAST likely prossibility and
using that as a scientific level of confidence
(in the results) is like looking at water half
way of a glass and saying it's half empty
instead of half full. And, with the amount
of genetic material that is looked at, combined
with small amount of samples for comparison,
taking 10 times least likely from the MLE is a
WEE BIT MUCH, IMHO (In My Humble Opinion).
But, to give you credit, I must say, based on
my last sentence, I can now see why your faith
is about zero in the NA & EA issue because of
the smallness of the samples to compare to in
both amount and different subsections of both
races. But, for the IE and the S-SA, I think
your faith (or lack thereof) is a wee bit pessimistic.
I can go into further detail to prove to you that
I understand this, but I don't think it'll be necessary.
But then again, I DO see the glass as half full,
so I may be wrong! LOL

Warm Regards from Maine,
Debby Peare

PS: I was wrong about the 10% ERROR.
As I understand this now, the lab tech from
DNAPrint was talking about the 2, 5 & 10 times
less likely factor, and NOT the error rate.
So, the error rate still stands at 5% (I think
you said Ray) and the test isn't worse!

This thread: