GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2001-07 > 0994009743


From: "Peter A. Kincaid" <>
Subject: Re: [DNA] Kincaid Surname DNA Project Plan
Date: Sun, 01 Jul 2001 14:49:03 -0300
References: <3.0.6.32.20010701010539.007aca80@pop.nbnet.nb.ca><3B3E3896.523B8938@wizzards.net><3.0.6.32.20010630145633.007b1c70@pop.nbnet.nb.ca>
In-Reply-To: <4.1.20010630223623.0b4a5690@wells.org>


>I personally do not have time to deal with excessive concerns about privacy
>or paranoia. I would rather these people just not participate in the
>study. The consent forms being used detail explicitly that under no
>circumstances will the DNA samples or results be released to anyone or
>organization without the express written consent of the participant. That
>is enough for me.


I hear what you are saying. Ultimately if someone is concerned
about a non paternity event or their DNA being made somewhat
public then they shouldn't participate. However, I have one point
to make.

Most of the don't care people step forward right away and offer to
participate with no strings attached. I suspect that most of you who
have put put this call out have gotten at best 10 to 20 participants.
Going back to my old University days 10 to 20 people doesn't give you
much from a statistical point of view. What we are doing is
sampling a total population of a particular surname. A true picture
of the population will be obtained by sampling about 1000 people
(this will give you a plus or minus a couple of percent). You won't
get anything really reflective unless you sample at least somewhere around
100 people (I forget but I think this might give you something like
plus or minus 15-20 %). Thus 10 people will only tell you whether
or not two of the participants have a common paternal ancestor within 30
generations. It tells you very little about what the true patriarch (ie.
males
directly descended from the founder of a particular surname) line is! Nor
can one say which of the two samples is a legitimate line (ie. an American
Kincaid patriarch could be the real father of one of his wife's seven
children - she having been impregnated by a hired hand the other six times).
Statistically possible!

Thus, if you want a good sample to determine a "surname or patriarch line"
my feeling is that you have to reach out to those who are concerned about
privacy to get your sample size up. Some surnames are quite common and
may not have this problem at all but for an uncommon surname like mine I
believe that this project may just waste a lot of money unless a good number
of participants is obtained.

I want to clarify that this is a different issue than the recent common
ancestor one. The latter is addressed by the number of markers
being tested (12 vs 15).

Best wishes!



Peter A. Kincaid
Fredericton, NB, Canada

This thread: