What Do You Do When the DNA Results Seem to be Lying?

I received an email message today that is not terribly unusual. I have received a number of similar questions before. I did reply in email, but I thought I would also write an article about it as I am sure others have faced the same “problem.” In fact, the resolution is simple, although a bit expensive.

This is the email message I received although I edited out the name of the person and the name of the DNA testing company in order to protect the privacy of both. In fact, this could have happened with any of the DNA testing companies:

I have a topic that has been bugging me lately. A certain DNA testing company is advertising about their “ethnicity” reports. My previous family history results show that I am over 80% British Isles and less than 5% German. However, I know that my father (he had his test done, too) is almost 50% German/Czech. Our family history research also shows that his father must have been close to 100% German.

I understand that I get what I get – not an exact % split of DNA but a roll of the dice. However, their commercials imply that you will know that you are not German if the DNA test shows no German in the ethnicity profile. What gives?

I think they are misleading people with those ads. What’s your opinion? (I also think their ethnicity reports are not 100% accurate.)

Thanks for your consideration of my question, and thanks for your newsletter. It has led me to several invaluable resources over the years of family records I would have never found otherwise.

Here is my (slightly edited) reply:

There are at least two possible reasons that the DNA results show non-German ancestry of the individual in question. The most obvious reasons are:

1. A mistake at the DNA lab where your test sample was accidentally swapped with a sample from someone else.

2. Did his ancestors REALLY come from Germany and nearby regions? Sure, that’s what the records show, but were those REALLY HIS ANCESTORS?

One “mystery” that turns out to be very common in DNA research is that someone was quietly adopted into a family some years ago without paperwork and other family members kept it quiet (this happened often; I have several examples in my own family tree).

Another possibility is what is humorously referred to as a “non-marital event.” That is, someone in your family tree spent at lest one night with someone other than his or her married and documented spouse. And, of course, there was an occasional case of rape.

While we all smile when we say “non-marital event,” the fact remains that such liaisons were common throughout the years, even in the 1700s or later.

There is a book called Sex in Middlesex by Roger Thompson. The book is not available online, but a well-written review of that book by Alicia Crane Williams may be found at: https://vita-brevis.org/2017/03/sex-in-middlesex/.

In the book, Roger Thompson describes various sex “crimes” that were tried in court in Middlesex County, Massachusetts, in the mid- to late-1600s. As reviewer Williams wrote:

“Sex in Middlesex pulls its facts from the Middlesex County, Massachusetts, court records. Eleven chapters discuss court cases by categories such as “Fornication: Detection and Evasion,” “Courtship and Patriarchal Authority,” “Pregnant Brides and Broken Promises,” “Unfaithful Wives,” “Unfaithful Husbands,” and “Community Control.” Statistical charts include “Geographical Incidence of Sexual Misdemeanors [1649–1699]” (the winner is Charlestown with 60, next was Cambridge with 31), and “Incidence of Conviction for Sexual Misdemeanors.”

I am sure such events were not limited to one county in Massachusetts. Anyone digging through old records can find many similar court cases everyplace else. While we commonly think of our ancestors as straight-laced Puritans or others who would never do such things, the fact remains that they were human beings with the same weaknesses and challenges that modern-day humans face. The court cases reveal that unmarried couples or couples who were not married to each other had extra-marital affairs probably at least as often as do today’s couples.

So, were your grandfather’s REAL ancestors from someplace other than Germany? Or did one of your ancestors have an extra-marital liaison that you do not know about? It is possible that the DNA results you have received may prove something about your grandfather’s ancestry that perhaps even he didn’t know.

Luckily, the solution is simple, although a bit expensive. Have another DNA test taken by another DNA lab. (I have had my DNA tested by four different DNA labs, and I know of other genealogists who have tested with even more labs than that.) See if all the test results agree.

If only one test result shows non-German ancestry, then the first assumption probably is correct: the lab made an error.

However, if all the DNA tests say that he had little to no German ancestry at all, then I would suggest that you have some new family history challenges ahead of you!

I can hear a collective gasp from everyone reading this article: “What? Not MY ancestors!”

But it was true many, many times.

So here’s a question for everyone else: What’s in YOUR DNA?

37 Comments

While I disagree that non-marital unions were as common then as today, they certainly did happen, and I’ve discovered at least two in my own family tree. The most interesting, actually involved a third cousin (DNA said he was a second cousin) who was adopted as an infant. Now, due to the wonderful advent of DNA genealogy, he has not only met me — his first revealed DNA Relative, but just in the last month he has tracked down four half-sisters, who were at first shocked to learn that their father had had a premarital liaison with a high school sweetheart. However, upon further testing, showing they matched on 25 percent of their autosomal DNA, they are conversing and plan to have an expanded “family reunion” in the near future.

Like

I love the comment, not in my family! Well guess what it is in MY family 🙂 and sometime that living person or their adult children will take an Ancestry test out of curiosity. Boy is that going to cause some Big questions! Can’t wait, can’t tell, if only I could 😉

Like

The results aren’t lying—the ancestral predictions cannot easily tell a German from a British person. At Ancestry that’s why they call them Estimates—if you dig a little deeper you will see that the area for Germany includes Southern England. The above article is misleading. Yes sometimes people aren’t who you thought they were but more often the problem is with the ability of the DNA to be that geographically specific. One example of a more responsible article https://www.legalgenealogist.com/2017/04/16/still-not-soup/

Liked by 1 person

    Wheatonwood you are right on with the article you refer to. It’s all about the databases the companies match to. Where do these companies find pure ethnic DNA for the baseline to match to????
    My test results have changed four times. You need to take DNA Tests results with a grain of salt.

    Like

    Thanks for the link. Very enlightening. I agree with this author’s view that DNA testing is for very different reasons, not determining your ethnic background, per se.

    Like

    The Saxons and their invasion … followed by the Normans and their invasion … all of it compounded by the old saw that “Mommy’s baby is Daddy’s maybe“.

    Like

    Germany wasn’t actually formed until around 1880, if you have 128 5 time ggrandparent, they could be from almost anywhere. Think of the bigness of families, and split your DNA, from everyone uphill, nonpaternal, is the dna term.

    Like

    To compound matters, they are mostly looking at the DNA of people who live in an area today which may not be the DNA make-up of the people who lived there when your ancestors lived in the same region. The ability to get decent amounts of autosomal from ancient remains is questionable at best and often can’t be done reliably if at all. If that’s not bad enough, the relative mobility of most areas tends to result in mixed DNA over time so your “German” may actually be British because the ancestors had German heritage, but wound up in England.

    Like

It is not an exact science no matter what the testing companies tell you. A lot of their results are based on other samples. The more samples they have, the better the results

Liked by 1 person

I had a grandchild who claimed our son was her father. He said this was possible. They did two separate paternity tests that came back negative. They were still sure of it, so they came to me. I advised them to do the Ancestry DNA test. He is half native, father is half Hawaiian/Japanese. When her test came back sure enough, 1/6 Native, 1/6 Japanese, and 1/6 Hawaiian. The mother being English which accounted for the other half. So what gives? Both parents at the time were using illicit drugs and a lot of them. Sometimes we here of mutations in genealogy. Of course the expert scientists will deny all this. In fact they have to deny it. Why? Can you imagine the lawsuits that would arise if they admitted this? The hard truth is we are still a couple of decades away from absolutely reliable test results.

Like

    Jeff, If two separate paternity tests came back negative, then your son is not the biological father of this child. You cannot base conclusions of parentage on “ethnicity” estimates. When you checked the AncestryDNA match list, were they listed as Parent/Child? That is what matters. Drug use would have absolutely no effect on a parent/child relationship being detected by a DNA test. Of course, experts/scientists will deny this possibility because it is simply not possible.

    Liked by 1 person

The most obvious explanation is that the customer has not studied the results closely enough to realize that the testing companies cannot easily tell the difference between British and German DNA. If in doubt, look at the match list to see where the matches’ ancestors came from. No need to jump to more alarmist conclusions.

Like

How about the fact that Germany (as well as France) does not allow DNA testing. It is disingenuous that ads do not say that. I believe that is why many are lumped into the British Isles. That said, do we really know the migration patterns of all of our ancestors? I doubt it.

Like

    LivingDNA is expanding into Germany for DNA testing now.
    I noticed in my 23andMe report, the percent for French and German had gone up from 7.2 to 7.9 while British went below 40 to 35.3
    So it is showing more and more are taking tests now, constantly refining my ancestry composition.
    Results from 3 DNA companies, yielded the clue to identity of my birth-2nd-great grandfather – French Canadian with some Native American mixed in since all my years of research I had NO known French Canadian and Native American except for this rapist because I was startled with many matches to French Canadians I recognized from my sister’s father’s ancestry.

    Like

This response is very misleading. None of the testing companies can accurately detect German DNA. The Germans, the French, the British and other north-western European populations are all genetically very similar and none of the currently available tests can distinguish between these populations. See this blog post “Is English DNA different from German DNA?

http://blog.ddowell.com/2015/05/a-year-ago-i-blogged-about-paucity-of.html

The writer compares his American wife’s DNA results from all three testing companies with my British results from the same companies. According to the testing companies I seem to be more German than his wife is! See also my blog post here where I provide links to useful resources:

https://cruwys.blogspot.co.uk/2015/05/comparing-admixture-results-from.html

If someone is trying to determine whether or not they have German ancestry they should be looking at their matches with genetic cousins to see if they match people with German surnames.

Liked by 3 people

elizabeth a ristau May 27, 2017 at 5:56 am

Not counting× DNA matches or non matches from recent generations, everything I have read suggests that the ethnicity results are from about a thousand years back. People have moved around and interpopulated a lot in that length of time. In Europe alone you have the Vikings, the 100 year war, and so on. The ethnicity part of these reports should be considered more as fun than as “whoa, what happened here?” DNA results are of course a different story when it comes to relating people more recently.

Like

Most of these current DNA companies are only testing AUTOSOMAL (middle relatives) which make up the major part of the specific database for that company. Yes, have more than one test done, if it SO IMPERATIVE! But have the testing done at the Y-DNA or mtDNA level. These do not come as cheaply as those purported. 2nd, consider that these percentages are only APPROXIMATIONS – not firm realities. 3rd: Get a good DNA book by Blaine Bettinger and hangout in the DNA blogs to learn more of the ins and outs insead of just complaining about understand the results. Carl

Like

My birth father was Armenian, with ancestors living in Harput, Turkey, and Beirut, Lebanon. My birth mother was French-Canadian and Swedish. I have definitively traced the vast majority of French-Canadian and Swedish lines back to the 1600s. My DNA shows 60% British (of which 20% is Irish), 30% Western and Northern Europe, 1% African, and 0% Middle East or Eastern Europe — the area where Armenians lived. Every far-fetched explanation for that simply doesn’t make sense — was my birth father REALLY Armenian — was I adopted (Yes, I said that in the beginning) — maybe my “Armenians” were really British, etc. The results are totally different from what I KNOW about my birth ancestry, so I have the feeling that I have wasted money on my DNA testing — until something more accurate comes along I shall not bother doing anything with the DNA. NO ONE has given me an explanation that makes any sense.

Like

    Michael, Do you have DNA matches to people with Armenian ancestry? That is the key. If you do not, then I would conclude that either your birth father is someone different than the one you know or he was not genetically related to his Armenian parent(s).

    Liked by 1 person

    Who did you test with, I would confirm that test with either Ancestry or FTDNA (whichever you didn’t use before) and upload both results to GEDmatch.
    An interesting note is that the largest population of Armenians outside of Armenia is in Glendale, Calif. I’m betting this US population has a pretty good sample population in the two largest DNA data banks. Just a guess in my part.

    Like

I have a project with more than 1000 members and have been tracking their ethnicity percentages since before and after the latest myOrigins update. As mentioned by the above posters, the algorithm struggles to differentiate between Northern Euro/Scandinavian and British Isles among other things. With the newest myOrigins we are seeing British Isles/Scandinavian where it had previously been WCEuropean and vice versa. Also, those who had not had Native American are now showing North and South American trace amounts. Some of those results are completely inexplicable on paper, particularly with families recently immigrated from the British Isles and no previous relation to the America within a genealogical time frame.

Like

I just had this conversation with a DNA company yesterday. The autosomol (sp) test does testing on different genes that other tests. My daughter has had 5 tests done and two turned out different The MtDNA was the same. The other ones were different. One was thousands of years in the making and the other was her own genetic makeup. One said she was NA and one said not. Mine was the same on both.

Like

Another possibility is as I read in the book, “The Invasion of Europe by the Barbarians” movement of tribes. One example is Lombardy, Italy. In this book it describes a German tribe, the Lombards, invading Italy and taking over that region. So in this case an individual may have ancestors from Lombardy but their DNA shows a heavy German DNA.

Like

JUDITH O. NEWMAN May 27, 2017 at 9:34 am

Look at the article on p 678 in the May 19, 2017 issue of Science (published by the American Association for the Advancement of Science) titled Bursting Myths of Origin.

Like

I understand that the ethnicity estimates of the various DNA tests usually do not reflect reflect where my ancestors may have resided in the past couple of hundred years. Instead, those estimates reflect the areas where the most distant of my ancestors resided – and those ancestors were probably living before many records were being kept. For example, my late husband’s immigrant ancestor, and two or three generations before him, were from northern Germany; our son’s Y-dna shows his “deep ancestors” to be from what is now Norway, Sweden or Finland. Those Vikings got around!

Like

As our 96 year old grandmother says….”You never know who slipped under the tipi flap!”

Liked by 1 person

Your article got me to thinking about the accuracy of my tree despite all the sources, citations, etc. We only know what we know. If any of our ancestors, male or female, had a “non-marital” event. Our tree at that point on is only half right but, it does get back to being more accurate as adittional generations kick in. If one of my great-grandparents is not who I think he/she is, I have a 1/8th error and dna would certainly pick that up, Further back, not so much. But, if, for example my paternal 5 times great-grandfather was not the father of my 4 times great grandfather, my surname may be correct legally and from a religious – baptism in my case – point of view, but, I suppose, a y DNA test of a relative with the same 6 times great grandfather, but a different 5 times great gandfather, would pick that up. I could actually have this testing done, but, I wonder if I want to know, because we wouldn’t know which of our ancestors went astray.

Like

“Lying” might apply to the ads but is unfair to the field of DNA ethnicity testing. This testing is a work in progress, with various difficulties:
— Trying to untangle many centuries of invasions and migrations, both large-scale movements of peoples and family-level moves.
— Limited DNA samples. Need to reconstruct the geography of ancient DNA from DNA of current volunteers. One careful study of UK DNA limited sampling to people whose four grandparents lived in the same locality. A good approach, but the map of the volunteers shows the unevenness of the sampling — large parts of Scotland with no data, Northern Ireland included but the Republic of Ireland not.
— Some ancient DNA has been extracted from archaeological remains, but the sampling is sparse, in both time and location. A recent study of the Beaker Bell culture had 170 DNA samples from ancient remains to cover Europe.
— Samples at DNA genealogy sites are biased in favor of people who are interested in high-tech genealogy and can afford the testing.
— Samples from the scientific literature are biased in favor of regions of interest to particular scientists studying of human migration. Sardinia and Orkney are well-studied.
— The sampling problem will lessen with time, as more people are tested for more markers.

One of the testing companies said I had a lot of Scandinavian ancestry. But my paper-trail genealogy has none. One explanation is that the Vikings or later Scandinavians invaded all my known ancestral lands — Ireland, England, West Prussia. Another is that the current methods of analysis cannot sort out the peoples of the North Sea very well. (Average ancestral location = Doggerland!) Both explanations may apply.

My suggestions: (1) Get tests from several companies. Convince your first cousins on both sides of the family to get tested. Comparing the results will give you an idea of how seriously to take the results. (2) Release your DNA data to researchers doing serious studies of DNA trees, following the Golden Rule for DNA genealogy: Be a data point unto others as you would have others be data points unto your tree.

Liked by 1 person

Suffice as to say, no one makes a cake recipe exactly the same every time.

Like

How about the case where the testing company’s “results” change over time? Three months ago I downloaded a screenshot of my mother’s DNA origins “map” from ftDNA showing her as 47% Scandinavian 9% southern Europe. Today I went to the same site, same company, same DNA sample and the map shows NO Scandinavian, 2% southern Europe. We have received no notification from the company why they have changed her result. What’s with that?

Like

I have had the same change from the well known autosomal testing company. I used to have “Scandanavia” as part of the ethnic mix but no longer. Given that I am 3/4 Irish (6 of my 8 great grandparents were either born in Ireland or had parents who were-the other two were French-Canadian (one born in Quebec, the other’s parents were born in Quebec.)
Yeah, there was a MAJOR piece of the discussion about Scandanavian influences- primarly “Vikings” but also the vast trade network of the Danes- in NW Europe and Ireland/England but now that has simply vanished from my ethnic mix.
I think they are having some challanges with their testing results and the database.

Like

As someone pointed out, “Estimate” is the key word. The population samples upon which ethnicity estimates are based are still relatively small.
Migration patterns and consolidation of early European tribes before nations existed result in mixtures of many kinds.
As I study microfilms of Catholic church baptismal records for the 18th and 19th centuries I find clusters of Illegitimate births. Further research usually shows that soldiers were in the area nine months earlier. Stuff happens.

Like

Good article Dick and some great comments ! It is alway great to learn something new and I like to try and distill my understanding down to it’s simplest components …
I am a novice newcomer to the DNA Genie scene, but I do think that it is very important to differentiate in the general use of terminology between ‘DNA results’ and ‘DNA Ethnicity Estimate’ reporting …
In my simple mind my DNA results should be identical from each lab, assuming that they all test the same areas of my DNA, do they ?? It will depend on the test that you pay for and I understand that most companies only test a portion of our DNA. My DNA results can then be used for matching to others with similar DNA results, the better the match the higher the possibility of a blood relative being found. This is an exact science and should never vary between companies if the exact same parts of my DNA are analysed in the labs.
My DNA ethnicity ESTIMATE however is a completely different scenario as it is NOT an exact science and open to interpretation. It will vary with each company depending on the DNA pool that they are matching my DNA against … and it may change over time as the reference pool changes and maybe a different method (algorithm) is used to perform this comparison. Ignoring the DNA companies hype, I look at this report as a best effort in an industry that I believe is still relatively in it’s infancy but learning exponentially.
We will always have a small part of our genealogical population that can not differentiate between the two and they will have far bigger expectations of the estimated reports than the current analysis can hope to achieve … no matter how many warnings the companies give that they are best estimates are … estimates !
The advances in DNA analysis over the last few years for me have been astounding, confirming some long held historical understanding, overturning other long held beliefs or finding new migration and settlement paths. It only seems like yesterday when I was riveted by Brian Sykes’ book Seven Daughters of Eve and sending my sample off to the first National Geographic DNA project Geno 1.0 for analysis … it was the dawn of a new era in our general understanding of our past … but it has not quite finished yet !
If my ethnicity reports change over time I can live with that … my DNA will never change … subject to Dick’s hint that there may always be a mistake made in labs getting samples mixed up 😉
David (London)

Like

I took the Ancestry DNA test and was surprised when it came back with the greater percentage Scandinavian. I always considered myself primarily German. My paternal grandfather was 100% German and my paternal grandmother also had German in her ancestry. The test didn’t indicate any German and the other highest percentage was European. Do they actually specify “German” on the test or is that part of the European?
Definitely confused. I also have a friend who considered himself English and his came back with more Scandinavian also. Maybe his wife with a Scandinavian maiden name will have more English in her DNA test? I didn’t question this before I read this article, but now I wonder what is up?

Like

    Take a look at history. Sweden invaded Poland in the 17th century and is is likely that a number of Swedes remained in that part of the world and migrated around including Germany. Likewise, Vikings repeatedly invaded the British Isles and created many cities.
    One article i read said that many of today’s Brits and Scots may have more Viking DNA than any other.

    Like

    Thanks, I thought of that too. I like the idea of having Viking blood!

    Like

My favorite is, “the dNA tests are wrong because my grandmother told me my gr gr grandmother was a Cherokee Princess”…Smile. Also, look at the surrounding countries. For instance, there is a small region in upper Italy which has been “ruled” by several countries, and each new ruler decided who could live or who had to leave, so the people who lived there would change their surname to “fit” the new authority. This included Italy under Mussolini”s rule.

Like

Leave a Reply

Name and email address are required. Your email address will not be published.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

You may use these HTML tags and attributes:

<a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <pre> <q cite=""> <s> <strike> <strong> 

%d bloggers like this: