GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2003-09 > 1062706796


From: (Raymond Whritenour)
Subject: Re: From Tony Frudakis RE: [DNA] Cecelia's Mother's DNAPrint 2.0Results
Date: Thu, 4 Sep 2003 16:19:59 -0400 (EDT)
In-Reply-To: DNACousins@aol.com's message of Tue, 2 Sep 2003 10:33:30 EDT


Ann:

You're much too kind to DNAPrint genomics, Inc. Nobody said that they
didn't explain their confidence intervals in the User Manual. In fact,
the User Manual tells us that the "best answer" for minor ancestry is
within the tenfold confidence interval. But... that's NOT what Dr.
Frudakis has been saying, in his correspondence on this list. Just this
past May, in a message to David Faux, he stated that the MLE is
"accurate" within a few percentage points, for minor ancestry--and that
anyone who didn't understand that doesn't understand the mathematics and
science behind the test. I would submit, to you, that a NA result of
13% (MLE) with a twofold confidence interval ranging from about -4% to
+30%, makes one of Dr. Frudakis's statements untrue. This twofold
confidence interval cannot be a "better answer" than the MLE, if the MLE
is only off by "a few percentage points" of the true minor ancestry, as
he stated in May.

Ray Whritenour

This thread: