GENEALOGY-DNA-L Archives
Archiver > GENEALOGY-DNA > 2008-12 > 1228465854
From: Beth Long <>
Subject: Re: [DNA] Data from related people (was Iberian R1b age estimates)
Date: Fri, 5 Dec 2008 00:30:54 -0800 (PST)
In-Reply-To: <813557.4778.qm@web111414.mail.gq1.yahoo.com>
Sorry, I should have mentioned that Rebekah Canada has also been an enormous help with our project.
Otherwise, I'm surprised that I get very little feedback (though I know our project data gets mined extensively).
Beth
--- On Thu, 12/4/08, Beth Long <> wrote:
From: Beth Long <>
Subject: Re: [DNA] Data from related people (was Iberian R1b age estimates)
To:
Date: Thursday, December 4, 2008, 11:51 PM
Re: Using data from related people
It used to be that our project (now 147 members) contained only those which
were either "clearly related", and those which were "clearly
not related". However, of late we have developed some clusters (or whatever
you wish to call them) in which the people appear to have a common ancestor, but
maybe not too recently. I must say that the mutation-colorized chart on the
Group Administrator Page is a great helping identifying these; it clearly
highlights the matching and unusual values among the
"cluster" members.
Such clusters are found among our members from haplogroups E3b and G in
particular. To me, it suggests members of the same culture who arrived in the
Carpathian Basin together a rather long time ago.
Actually, I have a lot of interesting data on my project excel chart, but only
Anatole has taken the trouble to help me analyze any of it.
Beth Long
--- On Thu, 12/4/08, John Chandler <> wrote:
From: John Chandler <>
Subject: Re: [DNA] Iberian R1b age estimates.
To:
Date: Thursday, December 4, 2008, 6:23 PM
John wrote:
> I think John C once said "more data is better than less data".
Yes, I'm sure I've said something like that, except that I would
have said "fewer" instead of "less".
> If your random selection from a family
> subgroup happened to pick the modal for the subgroup, and you know that
> there was an extreme outlier in the group, should there be an increase in
> the "confidence margins" for the calculations based on the fact
that by
> reducing the data used by random culling, you have increased the
uncertainty
> of the result?
That's true to some extent, but note that there is a *very* strong
correlation between the results for closely related persons. It is
usually very convenient in data processing to assume that all
measurements are independent and uncorrelated. If you use all of
the data and fail to account for the correlations, you will be
nominally reducing the statistical uncertainty, but the reduction
will be at least partly illusory.
John Chandler
-------------------------------
To unsubscribe from the list, please send an email to
with the word 'unsubscribe' without
the quotes in the subject and the body of the message
-------------------------------
To unsubscribe from the list, please send an email to
with the word 'unsubscribe' without
the quotes in the subject and the body of the message
This thread: