Go Back   Family Tree DNA Forums > Universal Lineage Testing (Autosomal DNA) > myOrigins Basics

Reply
 
Thread Tools Display Modes
  #1  
Old 2nd May 2018, 10:13 AM
mengel419 mengel419 is offline
FTDNA Customer
 
Join Date: May 2018
Posts: 2
REALLY unexpected ethnicity results

My sister-in-law's results were partly as expected: 21% English (her father's line), 33% Polish (her mother's). This matches my own research into her (my wife's) family that goes back four generations or more. Imagine our surprise when we saw 46% western/central European, which the map centered around Switzerland, southern Germany, and eastern France. That's huge. But no such origins anywhere in our records. A professional genealogist who has an open blog said "I really can't explain that." Can any of you?
Reply With Quote
  #2  
Old 2nd May 2018, 11:00 AM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,541
Quote:
Originally Posted by mengel419 View Post
My sister-in-law's results were partly as expected: 21% English (her father's line), 33% Polish (her mother's). This matches my own research into her (my wife's) family that goes back four generations or more. Imagine our surprise when we saw 46% western/central European, which the map centered around Switzerland, southern Germany, and eastern France. That's huge. But no such origins anywhere in our records. A professional genealogist who has an open blog said "I really can't explain that." Can any of you?
Dna results usually go back further than known family history. The result is not very surprising There was much contact between central Europe and Poland, e.g. Germans in Silesia and Prussia.
Reply With Quote
  #3  
Old 2nd May 2018, 12:03 PM
Frederator Frederator is offline
FTDNA Customer
 
Join Date: Jul 2010
Posts: 819
Quote:
Originally Posted by mengel419 View Post
My sister-in-law's results were partly as expected: 21% English (her father's line), 33% Polish (her mother's). This matches my own research into her (my wife's) family that goes back four generations or more. Imagine our surprise when we saw 46% western/central European, which the map centered around Switzerland, southern Germany, and eastern France. That's huge. But no such origins anywhere in our records. A professional genealogist who has an open blog said "I really can't explain that." Can any of you?
I'd be more interested if there were any significant patterns in her match list.

Ethnicity calculators are very poor at dealing with people of mixed heritage. It's just the trade off between relaxing matching criteria from the genealogical level of specific people to the anthropological level of aggregate populations--of course you're going to get all sorts of false positives.

It won't take you very long scanning through the posts on these forums to conclude that ethnicity calculators are just plain unreliable. You might find that some alternative reference populations and aggregation platforms at Gedmatch return results more consistent with the paper trail than others, for you personally, but you'll never find one single calculator that does it all for everyone.

This inability to reliably reproduce results for randomly selected people renders the significance of this whole enterprise as not much more than one of those 1970s dime store mood rings.
Reply With Quote
  #4  
Old 2nd May 2018, 01:11 PM
dna dna is offline
FTDNA Customer
 
Join Date: Aug 2014
Posts: 2,982
I'm glad that you did not pay for the services of that professional genealogist.

Are the results a little bit surprising? Only possibly, as it is difficult to say anything without knowing family names and places of birth (regions) of the oldest known ancestors.

In addition to various populations being present (see above posts), DNA of some ancestors might be disproportionately larger (and complementary slower) than simple arithmetic would indicate. That comes from the autosomal DNA inheritance process.


Mr. W.


P.S.
I understand that your wife DNA was not tested.
Reply With Quote
  #5  
Old 2nd May 2018, 03:25 PM
mengel419 mengel419 is offline
FTDNA Customer
 
Join Date: May 2018
Posts: 2
Thanks so far. No I didn't pay the genealogist, but she has quite a resume. I was under the impression that the testing only goes back five generations--older DNA wouldn't show up that prominently, would it? As for the "dime store" comment, in fact the results for my own FTDNA test were right on target. I understand the migration argument, but having traced back both families at least that far--the English side back to the 1500s and the eastern Polish to the early 1800s--I don't think it applies here. In fact, according to a blood relative who did a Y-DNA test, the English side probably stemmed from Scandinavia. Further comments and corrections most welcome!
Reply With Quote
  #6  
Old 2nd May 2018, 04:01 PM
DRNewcomb DRNewcomb is offline
FTDNA Customer
 
Join Date: Mar 2018
Location: MS Gulf Coast
Posts: 86
Quote:
Originally Posted by mengel419 View Post
..... "I really can't explain that." Can any of you?
My myOrigins breakdown came in at 60% British Isles (OK), 22% Swedish (OK) and 14% Southeast European (i.e. Italy & Greece). No German or Central European. My grandfather was born in an isolated little village in the Swabian Alps of southern Germany. I've traced his ancestry back to the 1700s. It's all Swabian.

My explanation of this, for the benefit of my relatives, is that while the science of analyzing a DNA sample is exact, the statistics of interpreting its meaning is not. You can feed raw DNA data into 12 different ethnicity models and get out 12 different results.
Reply With Quote
  #7  
Old 2nd May 2018, 04:03 PM
John McCoy John McCoy is offline
FTDNA Customer
 
Join Date: Nov 2013
Posts: 712
There may be some confusion about what can be detected by different tests.

The "five generations" threshold MAY refer to the limit at which actual cousins can be reliably detected as autosomal DNA matches: Depending on which set of statistics you look at, something like 50 percent of actual 4th cousins do not match each other using the normal autosomal match criteria. Around 10 percent of 3rd cousins also fail to match each other. I have some examples of this phenomenon ampng my own matches.

This effect is somewhat related to the problem of estimating the "ethnic" makeup of an individual based on admixture algorithms. Quite apart from the huge problem of constructing and verifying "reference groups" that use modern DNA samples to represent a hypothetical historical population, the fact that recombination and the random assortment of chromosomes operate to shuffle the DNA must affect the results. While you must have inherited 50 percent of the DNA of each of your parents, there is no guarantee which half you inherited. You could easily end up with far more than 50 percent of your father's paternal DNA, and correspondingly less of your father's maternal DNA, for example. After a few generations, this effect can compound to skew your actual "ethnic" DNA percentages so that they may not bear much resemblance to the "ethnic" percentages you would calculate from your pedigree. In other words, your DNA need not reflect your genealogy very closely at all.
Reply With Quote
  #8  
Old 2nd May 2018, 04:53 PM
dna dna is offline
FTDNA Customer
 
Join Date: Aug 2014
Posts: 2,982
Quote:
Originally Posted by mengel419 View Post
Thanks so far. No I didn't pay the genealogist, but she has quite a resume. I was under the impression that the testing only goes back five generations--older DNA wouldn't show up that prominently, would it? As for the "dime store" comment, in fact the results for my own FTDNA test were right on target. I understand the migration argument, but having traced back both families at least that far--the English side back to the 1500s and the eastern Polish to the early 1800s--I don't think it applies here. In fact, according to a blood relative who did a Y-DNA test, the English side probably stemmed from Scandinavia. Further comments and corrections most welcome!
When looking at the map of modern Poland (for example in Google maps), do ancestors of your wife and your sister-in-law come from the area delineated by the cities Kraków, Nowy Sącz, Krosno, Przemyśl, Rzeszów? Or possibly south of it?


Mr. W.
Reply With Quote
  #9  
Old 2nd May 2018, 09:11 PM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,541
Quote:
Originally Posted by mengel419 View Post
Thanks so far. No I didn't pay the genealogist, but she has quite a resume. I was under the impression that the testing only goes back five generations--older DNA wouldn't show up that prominently, would it? As for the "dime store" comment, in fact the results for my own FTDNA test were right on target. I understand the migration argument, but having traced back both families at least that far--the English side back to the 1500s and the eastern Polish to the early 1800s--I don't think it applies here. In fact, according to a blood relative who did a Y-DNA test, the English side probably stemmed from Scandinavia. Further comments and corrections most welcome!
The 5 generation limit applies to matches. Ethnicity estimates can go back 10,000 years depending on the source for the ethnic sample.
Reply With Quote
  #10  
Old 3rd May 2018, 09:23 PM
bhemph bhemph is offline
FTDNA Customer
 
Join Date: Feb 2018
Posts: 11
Let's say in the year 1 you have a dozen 1st cousins living in a village near Hamburg, Germany. A couple hundred years later some descendants have joined the Roman Legion as mercenaries. A couple hundred years later another group of descendants has joined with the Franks to invade Gaul, while another group was pushed another direction by the Huns. A couple hundred years later some join the Saxons invading England, while others move to the areas that were opened up by the Saxons moving. A couple hundred years later some descendants move east to avoid the forced Christianity conversions. Over then next few hundred years some are captured as slaves and sold in Moscow, Kiev, Oslo, Stockholm, Copenhagen, and Istanbul. Around the year 1000 some have moved to Bohemia and Poland working to Christianize the pagans, while others are moving after the Magyar invasions, and yet others still are in Hamburg as they have been for the last 1000 years. Some join in the Baltic crusades to help spread Christianity to Latvia, Lithuania, and Estonia a few hundred years later. More wars and displacement happen as the years pass and the borders we know come into existence. So now those genes that those original 12 first cousins had are spread all across Europe. How do you tell the group that went to England from the group that stayed behind? Or the group that went to France? Or the Baltic?

If your reference population is from the year 1, you know that they were in this area around that time from archeology. If your reference population is from the year 800, are they some of the migrators or some of those that stayed home and where do you classify them? If your reference population is a group of villages that say they haven't married people outside the area for 500 years and have documentation of it, what about where their ancestors were living 1500 years before the documentation ends? People moved around more than you think of at times. Even if you say each generation only moves 10 kilometers from their childhood home, after 100 generations they can be a long ways from where the family started. So there are a lot of possible pitfalls with ethnicity. Sometimes the selection made of where to classify some either or percentages are put where the user expects and sometimes not.

The Y-DNA possibly being from Scandinavia is only one of the how many millions of great x times grandparents living what 1000 years ago, 2000 years ago, 3000 years ago. The amount of shared DNA to any given ancestor decreases with each generation back, as the number of ancestors doubles. Eventually you run into the lack of population in an area and the same person is in multiple places in your ancestry, which can make it increase again but by that point there are probably some no DNA passed on ancestors as well since the amounts shared have grown so small by having gone back so many generations.

The specific algorithms used for many of the ethnicity calculations on different sites are proprietary and not shared. So it is kind of if you don't like the answer at one place, check to see what another has. Keep repeating until one comes in as a better match to what you expect.
Reply With Quote
Reply

Bookmarks


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Unexpected results LeahD Family Finder Basics 5 30th October 2016 05:14 PM
Unexpected results. Ann - Marie DNA and Genealogy for Beginners 1 13th January 2014 09:27 AM
Unexpected Results elliswr Paternal Lineage (Y-DNA STR) Advanced 4 16th June 2012 05:48 AM
Unexpected Results fcday Paternal Lineage (Y-DNA STR) Advanced 15 3rd April 2012 03:51 AM
unexpected results viofam Group Administrators - Advanced Chat 2 18th June 2005 01:14 PM


All times are GMT -5. The time now is 01:49 AM.


Family Tree DNA - World Headquarters

1445 North Loop West, Suite 820
Houston, Texas 77008, USA

Phone: (713) 868-1438 | Fax: (832) 201-7147
Copyright 2001-2010 Genealogy by Genetics, Ltd.
Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.