GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2006-01 > 1138736137


From: (John Chandler)
Subject: Re: [DNA] Newsweek Article: In your Blood (DNA Testing)
Date: Tue, 31 Jan 2006 14:35:37 -0500 (EST)
References: <3796-43DF8177-4009@storefull-3337.bay.webtv.net> <76ab240083539362be0720aed6ee2c8e@midcoast.com>
In-Reply-To: <76ab240083539362be0720aed6ee2c8e@midcoast.com> (message fromthomas goulde on Tue, 31 Jan 2006 11:02:25 -0500)


Tom wrote:
> In this instance a BGA told Mr. Joseph something quite different that
> what his private and social mirrors had been affirming throughout his
> life. His evident shock at the results clearly indicates the strength
> of his theretofore unquestioned assumptions - one finds in a mirror
> only that which one expects to see, is conditioned to see, chooses to
> see.

The real problem here is the marketing hype that goes with the test,
combined with a logistical drawback of the original form of the
triangle plot (which in the old days showed just the three ethnicities
selected by the quick-and-dirty method for reporting). If, as I
suspect, he was hit by the "zero African" bug, then he not only
received an arbritary and unintended "0%" report for the African
admixture, but also didn't get a picture of the confidence bands for
the African portion of his ethnicity. A reanalysis of his test after
correcting the bug would probably have shown up a small-but-nonzero
value, but even that doesn't really matter because an honost appraisal
of the "mirror test" is within the statistical margin of error of
ZERO. I would have guessed perhaps 10% African or less from looking
at his picture on the web. In other words, his "evident shock" was
based on poor communication of what the results meant.

John Chandler

This thread: