Saturday, November 10, 2018

The Amalgamation of Me: DNA Ethnicity Estimates

I'm a zealous convert when it comes to incorporating DNA into traditional genealogy. As a tool, it empowers research in ways never before imagined.

I speak from experience.

Genetic genealogy confirmed a long-suspected Non-Paternal-Event (NPE) in my own family tree (that four-part saga begins with A Family History Mystery Revealed). 

DNA connected me to my biological patrilineal pedigree, bypassing a fiction that had been touted as truth and perpetuated for more than 80 years. Without the insights from genetic cousin matching, it's doubtful that the truth would have ever surfaced or been confirmed through conventional paper trail research. 

Family historians imperil their own research when neglecting genetic testing.

But that's not what prompts many people to test

Let's face it, many consumers are not interested in - or even aware of - the full potential of DNA testing. Instead, they're lured to spit in tubes by flashy promotions that distill ancestral genetic testing down to simple geographical ethnicity estimates.

Why bother trawling through vital records, census enumerations, or tax and land deeds if you can get answers from salivating? Just wait a couple weeks and let the lab tell you who you are and where your people come from.

And I get it. To the average Joe, the appeal of this aspect of DNA testing fits with how we often casually discuss ancestry. 

"Where do your people come from?"

Which is met with an Atlas-grab of countries peppered with eyebrow-raising stories of a Mayflower voyage to the Americas and, for good measure, lineage anchored to [insert royal monarch here].

Of course, the DNA estimates are only as accurate as the sample populations against which your saliva is compared. They are interesting, for sure, but not something that has served my researched genealogy in meaningful ways. Not yet, anyway.

Don't get me wrong, I'm fine with the glossy appeal of ethnicity mapping features. After all, it's probably why millions of folks have tested who wouldn't have otherwise. Who knows, maybe one among them will help me bust through a research brick wall.

While I've long disregarded the ethnicity mapping featured by the companies with which I've tested, the recent results from a fourth company finally piqued my curiosity. 

Where do my people come from?

I've taken autosomal DNA tests with four of the leading genetic genealogy providers. How did my results stack up against each testing company? 

The results, while interesting, highlight the varying interpretations a tester is apt to get.

  • England, Wales & Northwestern Europe: 69%
  • Ireland and Scotland: 21%
  • Greece and the Balkans: 3%
  • Sweden: 2%
  • Norway: 2%
  • Italy: 1%
  • Portugal: 1%
  • Cameroon, Congo, and Southern Bantu Peoples: 1%

Family Tree DNA

  • West and Central Europe: 59%
  • British Isles: 39%
  • Finland: < 1%
  • West Africa: < 2%


  • British & Irish: 53.9%
  • French & German: 11.2%
  • Italian: 4%
  • Scandinavian: 3.3%
  • Iberian: 1.9%
  • Balkan: 1.8%
  • Sardinian: 0.2%
  • Finnish: 0.1%
  • Broadly Northwestern European: 18.3%
  • Broadly Southern European: 2.2%
  • Broadly European: 1.7%
  • West African 0.7%
  • North African & Arabian: 0.3%
  • Southeast Asian: 0.1%
  • Broadly East Asian & Native American: 0.1%


  • English: 31.7%
  • Scandinavian: 26.7%
  • Irish, Scottish, and Welsh: 21.9%
  • Iberian: 14.9%
  • Italian: 4.8%

There's a lot of variation across the European continent, which is likely the result of each company's differing sample populations against which my DNA is compared as well as their proprietary behind-the-scenes numbers-crunching. 

I think the results are clear: I'm an amalgamation of Europe with consensus on strong concentrations in Britain (God save the Queen!) and Ireland. My Italian ancestry is interpreted in varying fashions, and the Scandinavian consistently finds its way into the tally. 

Taken on the whole, I see where my paper trail dovetails with these estimates. But my research path isn't directed by these maps.

How do your estimates compare across companies? In what ways do ethnicity estimates inform your genealogy?


  1. Mine all show, quite accurately, that I am somewhere between 95 and 100% Ashkenazi Jewish. I needed no test to know that! But that wasn't why I took the DNA tests and had my mother, brother, and several cousins do the same. It was to try and break the brick wall on my mother's maternal side, and for that, alas, it has been pretty much useless except to confirm what I already knew: my mother is my mother, my brother is my brother, and the cousins are my cousins. DNA is not necessarily a useful tool for everyone's genealogy research. For an endogamous population like Ashkenazi Jews where we can get thousands of matches including over a thousand so-called second to third cousins who are NOT second to third cousins, it can be more a distraction than a help.

    1. You make a good point, Amy, and I wonder what the companies are doing - if anything - to try and account for endogamy. Perhaps science will one day be able to control for it and yield better, more telling information.

  2. Thanks, Michael---me, too. Some of the companies claim to adjust their match lists to account for endogamy. But if they are, I can only imagine how many matches I would have if they didn't! It is indeed very frustrating. I spent hours in the past week contacting and communicating with ten matches who all match my mother on the same segment with a fairly large shared (and triangulated) segment. We can't find any surnames or places in common aside from eastern Poland/western Ukraine---a rather huge area! So it does get me frustrated.... Sorry to vent on your blog!