Archiver > GENEALOGY-DNA > 2010-03 > 1269832143

Subject: Re: [DNA] 23andme cousin results
Date: Sun, 28 Mar 2010 20:12:50 -0700
References: <><><><><00d901cace95$a4b8bda0$5e82af48@Ken1><><00f401cace9b$2b915d00$5e82af48@Ken1>
In-Reply-To: <><><><><00d901cace95$a4b8bda0$5e82af48@Ken1><><00f401cace9b$2b915d00$5e82af48@Ken1>

Isn't there a statistician in the house who can come up with confidence intervals based upon some reasonable assumptions?

----- Original Message -----
From: Ken Nordtvedt<mailto:>
To: <mailto:>
Sent: Sunday, March 28, 2010 10:21 AM
Subject: Re: [DNA] 23andme cousin results

----- Original Message -----
From: "David Faux" <<mailto:>>

> No, but I have never seen any sort of document of this nature with
> confidence intervals. The FTDNA charts give a good idea of expected
> values

We should not let lack of such a document existing stop us from clamoring
for same. Remember the Alamo

I "fear" the tail of the distributions toward more distant relationships
will suck up much of the probability? But I'm not positive about this.
Such tails would render the expected value less useful for establishing our

This thread: