Select Page

Understanding the survey

Thousands of autosomal DNA customers of Family Tree DNA got an email this past week from Bennett Greenspan, president of the Houston-based testing firm, asking them to participate in a survey designed to help clarify some of the results that people see in what’s called the Population Finder. Because many folks got a little confused by some of the questions in the survey, and because the survey is time-limited (everyone should try to respond as quickly as possible), this weekend The Legal Genealogist reverses the calendar: we’ll look at this DNA question today and I’ll inflict my family on you … er … discuss my family … tomorrow.

The Population Finder, shown here with my results depicted, is a tool that compares the autosomal DNA of the person tested to that of known population groups around the world to suggest the amounts of the person’s ancestry that came from one or more of seven different continental groups.1 As you can see here (click to enlarge), my results show a strong European background but with a pretty high margin of error.

So the aim here is to try to do better. The email explained that the purpose of the survey is “to more finely tune the break-down of percentages in our Population Finder test” and “to more accurately reflect the genetic diversity of countries from around the world … (and) to more accurately reflect the country or countries and regions of the world where DNA similar to yours is found.”2

And, it noted, “citizen scientist data, especially from genealogists, is at least as important as what has traditionally been done by the academic community and therefore the answers to the survey are very important to us and to the general genetic genealogy community.”3

To ensure that I understood the survey questions, I spoke with Bennett Greenspan yesterday and found out more about the survey. First and foremost, it’s aimed at locating people with both genealogically and geographically known heritage. He explained that while the best way to find out about — say — the unique genetic footprint of the Scottish population would be to go to a fairly remote area of Scotland and test 30 or 40 people at random, the next best way would be to find the same number of people or more who have four grandparents from Scotland — and, even better, four grandparents from the same area of Scotland — and review their results.

So the survey hopes to find statistically significant numbers of persons who’ve already tested who have ancestors from specific areas.

Beyond that (and why the survey is also interested in folks like me who are true mongrels with ancestors who are a mix of recent immigrants and American colonists), the survey hopes to identify candidates to compare to the specific-area folks to see if it’s possible to be more precise in the admixtures4 detected.

While the questions are looking to get specific information about our ancestors, the wording of the questions gave rise to some ambiguity about just what information was sought.

The first problem was that some questions refer to a grandparent — say, the “paternal grandfather (father’s father)” — and others refer to the “most distant paternal grandfather (father’s father).” In context, it isn’t entirely clear whether this refers to one person (my father’s father) or two (my father’s father and the most distant male ancestor I know of on my father’s father’s side).

Greenspan’s answer: it’s two people. In total, the survey is looking for information about eight different people: the paternal grandmother and grandfather; the maternal grandmother and grandfather; the most distant known great grandfather and great grandmother in the paternal line; and the most distant known great grandfather and great grandmother in the maternal line.

So the question “What country is your most distant paternal grandfather (father’s father) from?” calls for an answer about the most distant known great grandfather you have on your father’s father’s side.

The next problem was the phrasing of the question, asking what country the person was “from.” Did “from” in this context mean birthplace or place of longest residence or something else? Greenspan’s answer: it’s the birthplace. The survey makes it clear that the answer should be stated in terms of the modern map (“using modern day borders”) to ensure that the data is consistent.

That same terminology issue confused folks in the questions asking what city a grandparent was “from.” Here too, Greenspan said, the question is asking for birthplace.

But the “what city” question raised another issue — what about all the people who weren’t born in metropolitan areas? For example, if a grandparent was born in a tiny hamlet or village, or even in an otherwise unorganized area of an early American county? How should the question be answered for them?

The answer is to give as much information in as much detail as possible. If you know that your grandfather was born, as mine was, in the town of Bad Köstritz within the Greiz district of the German state of Thüringen, that’s the answer to give.

Then there was one additional problem: in the questions asking about the most distant known great grandfathers and great grandmothers, the instructions say to answer United States/Canada/Mexico and the like “only … if you have Native American Ancestry.” So what’s the right answer for those of us who can take our lines back to colonial America but haven’t yet jumped the Pond to a specific location in Europe or elsewhere?

According to Greenspan, the answer in that case is to select the option Unknown. It simply doesn’t help in this deep ancestral-original inquiry to worry about the DNA after it got here in America and started being mixed up in our Melting Pot.

So, for the guidance they may provide, here are my answers to the survey (parentheticals not included in my survey answers):

     • 1. Are all four of your grandparents of the same ethnic group?

            No. (Two were German-born, two were American-born but not of German heritage.)

     • 2. What country is your most distant paternal grandfather (father’s father) from?

            Germany. (My earliest known great grandfather on my father’s father’s side was from Sachsen-Anhalt.)

     • 3. What city is your paternal grandfather (father’s father) from?

            Bad Köstritz, Greiz, Thüringen, Germany.

     • 4. What country is your most distant paternal grandmother (father’s mother) from?

            Bremen, Germany. (All of my grandmother’s people that I’ve traced so far were from Bremen.)

     • 5. What city is your paternal grandmother (father’s mother) from?

            Bremen, Germany.

     • 6. What country is your most distant maternal grandfather (mother’s father) from?

            Unknown. (We trace back to colonial times but can’t say for sure where from there.)

     • 7. What city is your maternal grandfather (mother’s father) from?

            Iowa Park, Texas. (It is a city, with all of about 6,000 people. Look it up.)

     • 8. What country is your most distant maternal grandmother (mother’s mother) from?

            Unknown. (We trace back to colonial times here too but can’t say for sure where from there.)

     • 9. What city is your maternal grandmother (mother’s mother) from?

            Eagle Lake, Texas. (With all of 3,700 people, it too is a city. Really.)

And I answered 0 to all of the questions about how many of my great grandparents were Native American, African and Jewish.


  1. See generally “Understanding Results: Population Finder,” Family Tree DNA ( : accessed 28 Sep 2012).
  2. Bennett Greenspan, Family Tree DNA, to Judy G. Russell, e-mail, 25 Sep 2012, Ancestral Origin Survey; privately held by Russell.
  3. Ibid.
  4. Family Tree DNA defines “admixture” as “ancestry from more than one recent population group. Many people today have ancestry from more than one population and/or location.” Family Tree DNA, “FAQ: Glossary,” entry for “admixture” ( : accessed 28 Sep 2012).
Print Friendly, PDF & Email