GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2003-12 > 1072139254


From: David Faux <>
Subject: Re: [DNA] The new DNAPrint test
Date: Mon, 22 Dec 2003 16:27:38 -0800 (PST)
In-Reply-To: <08cd01c3c8e4$a8597700$d6ebb8a1@claytonpq95u52>


Patricia and all:

I trust that you have been following the postings over the past year about the DNAPrint Version 2.0. The bottom line is that for genealogy the test has great reliability (accuracy in detected the same allele value when administered on more than one occasion), but falls badly on validity - measuring what it is supposed to measure.

Upping the number of Deep Ancestral Informative Markers from 72 to 180 will only serve to tighten up the confidence intervals - unless they fiddle around with the agorithm. Their problems will not go away until they collect large representative samples from around the world to use as their reference base. The sampling problems have been discussed at length. We must all face the facts that we have the right to get excited because we hoped to find some Native American, because that is our right, not because the results of the test can in any way depended on. The only exceptions are percentages above about 30%, and where the result tallies very closely with a clear paper trail (as does mine - but I really think it was just a random finding). Please everyone, check or re-check the log of Charles Kerchner which will show expected versus observed findings from the DNAPrint test.

I am not trying to be a wet blanket, but there is not a scientist I know who would support the use of this test to explore the possibility of small percentages of minority ancestry. The results could mean anything - false positive, false negative, or as so frequently been seen, downright bizarre.

David F.

Patricia <> wrote:
I too have been told by Family Tree DNA that I will be getting the new test
done as I have been waiting so long for my results.
Looking forward to it.
Patricia





Dr. David K. Faux, P.O. Box 192, Seal Beach, CA, 90740, USA



www.davidkfaux.org



This thread: