Source:DNA Matches

From CoxGenealogy
Revision as of 22:18, 6 September 2016 by Seanmcox (talk | contribs) (→‎High Level Statistical Analysis: Clarified the verbage some.)
Jump to navigation Jump to search

The following data comes from YDNA testing of Ronald J Cox. For any data entered into this page, living individuals should not be referred to in a directly identifiable way unless they have given permission to appear on this wiki. (This should be the case if they already have a page.)

Wallen/Walden

The closest and most promising match of all comes from a set of lines that don't even have the same surname. All of these matches suggest a single common Cox ancestor. Possible scenarios include infidelity, premarital sex, early marriage followed by divorce and remarriage, or adoption. This connection appears to be at or just a little older than each line's oldest identified ancestor.

Line Dist. Marker 37 Dist. Marker 25 Dist. Marker 12 Most Distant Ancestor Notes
A 1 1 0 George Washington Walden/Wallen(Campbell Co TN?)
B 2 2 0 Evan Walden, b. 1776 and d. 1852
C 2 2 No datum Evan Walden b.1776 TN Match doesn't show up when I include marker 12 in search. Seems to be hidden.
D No datum No datum 0 Evan Walden b.1776 TN Seems to be an odd entry in the database. Can't get contact info. Can't see how many markers were tested. Possibly due to privacy settings.

A family tree is available for line B which shows (most-distant to least-distant):

Walden Family Line
Evan Walden, Born: Tennessee, USA 1776
George Washington Walden, Born: Tennessee, USA Jan. 29, 1811, Died: 1891
Giles Jasper Wallen, Born: Kentucky, USA Jan. 11, 1846, Died: Feb. 17, 1937
Newton Wallen, Born: Wayne, Kentucky, USA Jan. 15, 1870, Died: Paso Robles, San Luis Obispo, California, USA Dec. 16, 1958
Herman Hull Wallen, Born: Tennessee, USA Jun. 19, 1900, Died: Redlands, San Bernardino, California, USA Feb. 28, 1967
Subject B

This chart makes clear that line A is common with the other lines, even though it shows a different most-distant-ancestor.

Since it is not certain how the other lines connect or diverge, it would not seem possible to mix results to get refined statistics on where a common ancestor would be expected to be. Line B will be used for most statistical analysis.

The Cox family line is, for comparison, provided below:

Cox Family Line
J-L70
William Cox
John Cox
Daniel Root Cox
John Nelson Cox
Daniel Franklin Cox
Marvin Luther Cox
Ronald J Cox
Sean Marvin Cox

6 generations back from Ronald J Cox would be the earliest likely candidate, which gives us William Cox. 6 generations back from Subject B in the Walden family line brings us to the father of Evan Walden. We might be able to almost completely eliminate an earlier common ancestor if we knew that lines C or D descended from Evan Walden through a different son. (I don't know enough about him to know even whether he had more than one son.) So, it would seem that William Cox, or one of his near ancestors would be the most recent common ancestor.

High Level Statistical Analysis

Based on markers 37 and 12, we would estimate the distance to the most recent common ancestor would be about 6 or 7 generations, with a significantly smaller 90% confidence interval. However, taking marker 25 into account would seem to broaden the results a bit. Without doing any low-level analysis, I would estimate an expectation value of about 8 generations. (I hope to to some low level analysis in the future.)

Taking line A into consideration would suggest a smaller expectation value, but without more detail on the line, it would seem impossible to add this to my analysis in any very valuable way beyond this mere suggestion, so it will probably not be attempted.

We were able to control against the possibility that the common ancestor was a Walden by comparing Wallen/Walden matches with Cox matches. Wallen and Walden matches are all listed above, and all trace back to a common ancestor somewhere more recent than the expectation value for the most recent common ancestor for lines A and B. More particularly, each individual match can be shown to have a similar or more recent expectation value for the number of generations to the most recent common ancestor when compared with our line. This suggests that these Waldens form a distinct unrelated group from other Waldens. On the other hand, when comparing with other coxes, we have 2 of 16 matches that show a statistically significant likelihood that the most recent common ancestor is older than the one connecting us to the Waldens/Wallens. (Many of the other results may be older, but simply don't provide enough data to draw an obviously meaningful distinction along these lines.) This suggests that the Cox line is related to other Cox lines very generally. This result was confirmed by an interested party connected to line B, from their perspective. Even further, they had noted that they simply weren't in the same haplogroup as most Wallens/Waldens. Their line had appeared out of place.

Resources