How accurate are DNA Ethnicity Estimates?

 

In our complex society, we all live with estimates. An estimate is a guess regardless of whether or not it is called an estimate, a prediction, or a trend. Not too long ago my wife and I contracted for some remodeling work in our basement. We got an "estimate" of the cost from a licensed contractor for a certain amount. Invariably, the amount we ended up actually spending was significantly greater than the original estimate. 

As the large online DNA companies began doing business, they promoted their "products" using estimates of the ethnic origin of people around the world. By matching DNA testing results with a pool of people assumed to be from a particular geographic area of the world, they claimed to be able to match you to your origins. Theoretically, this is possible if you have a large enough pool of reference DNA tests and some way to assign the results to a particular geographic location. See my previous post, "MyHeritage and the evolution of Ethnicity Estimates."

The question is how large is large enough? Here is a copy of my Ethnicity Estimate from MyHeritage.com.

This estimate corresponds roughly with my own research over the past almost 40 years. Now, here is a screenshot of my estimate from Ancestry.com. 

There are some significant differences between these two "estimates." Why would there be a difference? This is an interesting question. Let's look at my eight Great-grandparents. All of them were born in the United States except one born in Denmark. If I look at the sixteen parents of each of these eight ancestors, I find the following:

• United States - 6
• Denmark - 2
• England - 2
• Ireland -1
• Wales -1

I could keep going back looking at the birthplaces of each of my direct line ancestors (those from whom I could have inherited DNA) and I would still find some ancestors who were born in America back to the early 1600s. Other than Denmark, my research back as far as the 1600s with some few validated lines into the 1500s, none of these direct line ancestors came from Norway, Sweden, Italy, or any other part of the world. Going back into speculative ancestry, there may be one of two lines that finally connect with continental Europe in the Netherlands and France.

What do the DNA test results tell me that I don't already know from my long years of research about my ethnicity? The key to answering this question involves issues with specific ancestral lines. I have had DNA results help to answer some of the more difficult genealogical issues we could not answer with paper research. The value of DNA testing from this standpoint is undisputed. But again, what about these ethnicity estimates?

As long as we view them as promotional or advertising, there is no problem at all. However, over time, the ability of these large companies to focus on more exact and specific geographical areas might help to support some more of my own research conclusions or help me to look at other options. But right now, telling me, for example, that I have "Mountain West Mormon Pioneers, is not really helpful for my own research. What this means is that DNA ethnicity estimates can suggest avenues for research and confirm research that has already been done, but it is not particularly useful in specifically identifying new ancestors unless the individual researcher has done an extensive amount of traditional research. 

For those people who obtain a DNA test and never make an effort to research a family tree, ethnicity estimates are merely entertaining and not really relevant. However, to the extent that a DNA test gives someone the incentive to begin actually doing some research, these tests and very useful. 

How accurate are DNA Ethnicity Estimates? Not yet particularly accurate but over time, they may become very useful tools for more and more people but only those people who become involved in researching their family lines and using one or all of the large online family tree websites.