How Accurate are DNA Kits used for Testing Ancestry?

dnaMillions of people are purchasing and using home DNA kits to determine their ancestry. The television program Inside Edition enlisted the help of two sets of identical triplets and one set of identical quadruplets to investigate the accuracy of the at-home tests. The ancestry of each group should be absolutely identical since they all came from the same egg.

Test kits from 23andMe, FamilyTree DNA, and AncestryDNA were used.

The results are surprising.

The medical experts were confused and disappointed because of the varied results. Does this indicate that some of our our beliefs about DNA are wrong?

You decide. You can watch the television program’s tests and the results at https://goo.gl/Kz5ODH.

29 Comments

Well they did not say that the test kits were bogus. 2 sets seemed to have differences, and 2 sets appeared to have identical results.

Like

They said that all the sets were “identical” but fraternal twins etc are more common than the identicals. How could they be sure they were actually supposed to be more similar than other siblings?

Like

it would have been more meaningful if all had done all three of the different companies…then the results would have had more credibility

Liked by 2 people

The ethnicity analysis that these people had was just one of the test results offered by DNA companies. This analysis just uses an algorithm to determine an ethnic estimate. I fear that this brief report might cast doubt on the validity of using DNA testing. The three companies also offer a matching service with other test participants that many use in conjunction with their family trees and genealogical research. I have tested with all 3 companies and have found unknown cousins using each one. DNA tests are incredible tools for finding ancestors who are impossible to find using historic written records. All 3 companies give you the same data – it’s how the data is interpreted and how you are using it that makes DNA testing valuable

Liked by 2 people

The way the results were reported was confusing to me. Were all the differences between siblings confined to the ethnicity estimates? Or did they find differences in the actual DNA itself?

Like

It’s important to bear in mind that “Inside Edition” is a tabloid with a history of manipulating situations and facts in order to get “good TV” out of it. I’m a former TV news producer and can say that the real story was probably pretty simple and easily explained, yet they chose to cut it out. It’s the model shows like that are built on, unfortunately. It’s why the rest of us despise them.

Liked by 1 person

    Sorry, I’ve seen similar results in friends who have had testing, so doubt that the show
    manipulated it. Different companies do differently qualities of work and it’s a issue. Lots of sales hype and not enough good science.

    Like

One of the set of triplets looked to me more like the usual set of twins plus a fraternal that is more common. That would explain some of their results.

Like

This report focused on the ethnicity estimates, which are known to be variable (although presumably not among identical siblings), but I want to mention that things can also go wrong with Y-DNA results (and so presumably with autosomal and mitochondrial tests also). I took a Y-DNA 67-marker test with FTDNA and found no close matches, and no matches at all beyond the 12 or 25 marker level. Disappointed, I asked my brother to test also, just as a cross-check that there was not some mistake. Surprisingly he had many more matches, although his Y-DNA should be the same. Also, he and I were shown as matches at certain marker levels and not at others! Looking closer at the detailed marker results I found that he and I had identical test results. I then emailed FTDNA and pointed out the inconsistency. They agreed it was some sort of error, and, after a few more communications, indicated it had something to do with an error in loading the test results onto the system or web site that produces the matches. They then fixed that, and now my brother and I show identical matches, and we match each other at all marker levels. Unfortunately, we still don’t have any close Y-DNA matches on FTDNA, their associated Y match site, or on the Spencer surname group. Sigh!

The bottom line is that the DNA results may be correct, but the reported matches may not be if some error occurs in transcribing the data within the company’s systems. If you are not getting matches or have some other reason to be suspicious it might be worthwhile to get an appropriate relative for the type of test you took to test as a cross-check. I am pretty sure my autosomal results are correct because several known distant relatives had already tested and showed up as matches.

Like

    My husband had 37 and then 67 markers yDNA tested in 2009—it was 2014 before we got a surname match at the 37 marker level, and March 2016 before he got a surname match at the 67 marker level, GD of 3. Have sent two emails to the fellow with no response—frustrating!

    Like

    Wow. My name also is David Spencer, I’ve done genealogy research for decades, but on my paternal line I’ve never been able to go back further than my gg grandfather who, in census records, is listed as being born in the early 1800s in Maine. I did the FTDNA Y-67 test in early 2012, and just like you, I matched no one at any marker level. (I also noticed, at that time, on FTDNA’s Facebook page that some people were questioning why newer batch numbers were being finished earlier than older batch numbers…curious.) I too have no matches on the Spencer surname site or at Ysearch.org. A year later I ordered their Family Finder kit (so they would test a different sample) and matched several family lines that I knew were descendants from my great grandfather’s brothers. Later in 2013 I ordered 23andMe’s test kit and the autosomal results also connected to some known paternal lines. So, like you, some autosomal results show ‘matches’, but the Y DNA test, nothing.

    Like

I wasted over $200 only to find out that, despite my Armenian birth father, I had NO Middle East DNA, and that although all of my traced ancestors are from Quebec back to France, or back to Sweden, I come back with 60% British. People have come up with the most ridiculous explanations for this, none of which make any sense. Sure, maybe my Armenian ancestors all came from England originally. Or, maybe I’m adopted (ah, yeah). I shall not waste another cent on DNA testing until I can get something more accurate than DNA guessing.

Like

    I hope you contacted the testing company’s help line to discuss the fact that your ethnicity report is so completely at odds with the origins documented in your extensive family tree. Perhaps they might want to do a retest to either confirm or correct the results, or take a closer look at the comparison between your sample and the reference population(s) to which it was compared.

    Liked by 1 person

    it is also entirely likely that they have mixed up your DNA material with another it can happen,…and so the inaccurate info

    Like

    I don’t know if you test through Ancestry, but the Great Britain category for them is also found in France.

    Like

    Armenians aren’t middle easterners, west asian/east european.

    Like

I guess the program-makers assumed that identical siblings have identical DNA. A number of studies have shown that there are small differences – see, for example, this 2008 article from Scientific American: https://www.scientificamerican.com/article/identical-twins-genes-are-not-identical/

Liked by 1 person

It is misleading to suggest poorly informed TV programs’ “gotcha” episodes can truly help with “you decide” when it comes to using DNA in genealogy.

As noted by others, these types of programs very, very rarely discuss cousin matching, which is of course the key for using DNA in genealogy.

And also missing from so many of these TV programs is an informed discussion of what is truly meant by “ethnicity” and why so many people have preconceptions that are in conflict about what we have discovered.

Liked by 1 person

    To be fair, the testing companies generally advertise their tests as ways to determine ethnicity, not as a means of finding cousins. (This is one of the reasons why an increasing number of people who test haven’t uploaded a tree – they’re not researchers like us.)

    Like

Why bother finding triplets or quads? Just test the same person at all the companies. Yes ther will be some variation — they test somewhat different markers and have different algorithms. Also, send in multiple tests from the same person to the same company. Triplets make for better TV, though.

Liked by 1 person

Fact is these DTC companies are unreliable. 23andMe collects DNA to sell to big pharma for big profits and gives the consumer exactly what the paid for…a meaningless piece of paper

Like

Linda is absolutely correct in stating that one person should be tested with all kits. Additionally, more than one person should be tested to get a statistically significant result.

My cousin and I used the kit from The Genographic Project to do our tests, as we both concluded that it was the most scientifically sound.

So Dick, would you consider starting this and order the other three kits that you have not used to determine your own results?

Like

Why didn’t they download the DNA raw data and compare the results. This should’ve been the first step in their “test”.

Like

I am interested, as is my young daughter (13YO), in getting her a DNA test for her birthday – a request of hers. Which of these would you recommend? MY genealogy runs pretty deep, as it is rooted with the Huguenots. But she wants to know more. Proud of my curious girl. THANK YOU.

Like

H-

For kids, 23andMe is a good option. It not only has ethnicity results, it is the only one that shows both pairs of 22 (plus X) chromosomes’ ethnicity. Though it is medical/health trait oriented, that’s a good thing should they take biology in the future. There are other features, like short videos on recombination, that can hold the attention of teenager much better than the family tree useful features that we genealogists desire. I’m thrilled by 15 year old niece has expressed an interest as I’ve already tested her parents (and her grandmother, aunt & uncle) on 23andMe. Her parents ethnicities are VERY different so I can’t wait for the results to show her what is passed down to her, that came from many generations before. The mitochondria DNA story, which you get from the same spit test from 23andMe should also be of interest to her.
Congrats on having a curious teenager !

Like

    Someone whose parents and other relatives have already been tested by one of the genealogy-oriented companies and who is more interested in scientific research than in genealogy, might want to consider testing with the National Geographic Genographic Project.
    ( See: https://genographic.nationalgeographic.com )
    This is the test that was used by the “Finding Your Roots” program. It’s a bit pricier and doesn’t provide matching services, but it is a real scientific project seeking to understand where the first human beings originated and how they came to populate the world. As always, be sure to read the small print before deciding.

    Like

I have found Ancestry DNA to be fairly if not 100% accurate. It has also found cousins of mine from 4-8 generations (so far back). If someone in your family bothered to write names, dates, etc. down you will find your family. The Ancestry DNA also matches you to other family members you did not know about. Give it a try.

Like

I just got my results back. My only brother had me do mine to compare to his. The report made me laugh. His came back 20% Irish and British. Mine came back 18% Italian. His had no Italian and I had 2% English, which I assume is British. We were both born on an Island to the same parents. He legally changed his first name to a common name, I kept my uncommon birth name. There’s an opera singer in Italy who shares my name. His name is common in most English speaking countries. Coincidence, maybe, or maybe not….. One of my in laws, whose mother was Native American, her parents were both Native American and his test results had zero percent Native American. He was named after his Irish father. His results came back 64 percent Irish. Something doesn’t seem right to what we all expected.

Like

Seems to me if they can pass off 99.9% on paternity the should be able to due a bit better on the rest of it. I know for a fact my daughter is mine. So did a DNA for paternity on my daughter, so there is no doubt. (Certified Lab) then an did ancesteral type DNA on myself and my daughter, wife had one peviously from same company and according to them, my “male” daughter is mostly from the Crimean area, although my wife and myself are 82% and 90%, respectively British Islands (Scottish), waiting on reply for false advertising litigation. Now thats special

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: