You may have DNA match relationship issues and not even know it! Confirmation bias may be affecting your judgment about how you’re related to your DNA matches. You’ll want to know the difference between a genealogical v. genetic relationship, too. Here’s why.
This article was updated in March 2020.
Confirmation bias and DNA matches
One of the biggest problems we have as genealogists, and especially as genetic genealogists, is in confirmation bias.
Confirmation bias: the tendency to process information by looking for, or interpreting, information that is consistent with one’s existing beliefs. -Encyclopedia Britannica
If we see an ancestor in the chart of a DNA match, and we are pretty sure this is also our ancestor, we might be tempted to say that your DNA connection with this match proves that this couple is your shared ancestral couple. But that’s not always the case. You can use the shared DNA to support a connection, given other genealogical information, but DNA can’t ever really prove a connection to anyone outside of your immediate family.
There are several things you can do to avoid confirmation bias when determining your relationships to your DNA matches.
- Check your match’s family tree to see if there are other ancestors you might have in common. (Doesn’t this just seem sensible?)
- Look closely at the genetic relationship possibilities indicated by the amount of shared DNA you have with your matches.
- Compare these possible genetic relationships with your proposed genealogical relationships to make sure they’re consistent, or at least they’re not wildly inconsistent. A genealogical relationship is one that appears on your family tree. It’s documented by family memory and traditional genealogical research. A genetic relationship between two people is circumscribed by the amount of DNA they share.
- Build genetic networks with your DNA matches. More on that below!
There are scenarios in which DNA doesn’t reveal an actual genealogical relationship. Let’s say that Meredith, a documented descendant of your 3x great grandparents Martha and Matthew, had her DNA tested. Your genealogical relationship to Meredith is 4th cousins. However, when the testing company completes the test, Meredith is not found on your match list! Does this mean she really isn’t related to Matthew and Martha? Not necessarily. Even though your genealogical relationship is 4th cousins, your genetic relationship can be non-existent. You will only share DNA with your 4th cousins about 50-60% of the time.
It is also possible to share a certain amount of DNA with someone who isn’t actually your cousin! You will share DNA with individuals who do not share a recent common ancestor with you, but who just share a similar heritage. For example, people from Italy may share DNA with each other not because of a single recent common ancestor, but because of a more distant connection to Italy.
Genetic networks and your DNA matches
Probably the most powerful thing you can do to understand your relationships to your DNA matches is create a series of genetic networks to gather more evidence. This involves using the shared matches tools on your DNA testing website to see who matches both you and your match.
To learn more about working with DNA matches, check out our AncestryDNA Tour. In addition to an overview of working with your matches, you’ll learn about the other features Ancestry offers and how you can make the most of them. The AncestryDNA Tour includes over 90 minutes of video instruction divided into 16 different segments, and comes with a printable, interactive workbook, so you can apply what you learn to your very own tree.
Tell me more about the AncestryDNA Tour!
The steps listed in this article constitute part of “reasonably exhaustive research,” which genealogy standards of scholarship say must be accomplished before you can place a DNA match on your family tree with any degree of certainty. Remember, DNA is just another record to use in your journey to document and tell the story of your ancestor, and it can be powerful, if you know how to use it.
An earlier version of this article appeared in December 2018 on genealogygems.com. Rewritten and updated in March 2020 for Your DNA Guide.
Thanks for all your information. I have been working on my family tree for about 35 years now. I found you while being angry because of all the sites, especially in Europe only care about YYYYYY. and when a woman has questions about the females of her line or about the genetic mtDNA testing our questions are ignored or put down, etc. Do you know of any family tree type sites that are only for women? This would be nice with women geneticists and genealogists mentoring it. Well thanks for letting me vent and thanks for your site. Victoria
Hello Victoria, thank you for sharing your frustrations, it sounds like you have had an A-typical experience in the genealogy community. We are quite an open and accepting bunch! While there aren’t any sites specifically just for women, there is a site working on creating a database and resources for mtDNA (and YDNA). Try http://www.mitoydna.org.
Hi Diahan
Thank you very much for the article and material, it is helping me think through a puzzle I have. The puzzle is that my DNA matches on my father’s side show 1. a 1st cousin once removed which is confirmed through obvious genealogy. 2. a match that is most likely a 2nd cousin or a half-1st cousin once removed with no obvious geneaolgical link. The part I was confused about is that these two matches are not shared matches but I think that is still ‘possible’ given your article.
Thanks, Nick.
Nick, this is a situation I would want to take a look at. I assume you know both matches are on your dad’s side because they do not match your mom, or another close relative on your mom’s side?
Two close matches like these on the same side of your tree should share DNA with each other. Essentially, everyone closer than 3rd cousins should shared DNA with each other. So perhaps these two matches need a deeper review.
Diahan, thanks for getting back to me. Yes, my mum has done a DNA test and neither of the two matches I have show as being on my mum’s side. For the person (Mr A) where we can’t establish a genealogical link, the predicted relationship is 63% 2nd cousin; 1st cousin 2x removed; Half 1st cousin 1x removed; Half great-grandaunt/granduncle; or Half great-grandniece/grandnephew. I am ~50-y-o and he is 72-y-o. For the person (Mr P) where there is a genealogical link (1st cousin once removed), the predicted relevant relationship is 53% 1st cousin once removed and 47% half 1st cousin once removed. I am now wondering if there is a possibility that I am linked to both grandparents and Mr P is linked to only one and Mr A is linked to the other one. This is the only scenario I can imagine to explain the links between me and both without a link between them.
Nick, it looks like your assessment may be correct. It looks like your most likely relationship could be half first cousin once removed. And that would explain your findings.
Hi, Diahan,
I enjoyed watching your virtual tips and presentations at the RootsTech Conference! I’m puzzled by something in Ancestry DNA matches that probably has a simple explanation. Ex., Conor and I are a 5th-8th cousin match, and we have a shared match with Joseph. Joseph and I are a 3rd-4th cousin match, but we do NOT have a shared match to Conor. This is not an isolated example. How does one interpret this? Hope you can help. Thanks. Cathy
Thanks for coming to RT Connect!
First of all, it is important to note that Ancestry will only show individuals as shared matches who are sharing at least 20 cM of DNA with BOTH of you.
But this doesn’t explain your situation…so the short answer is, I don’t know of an explanation for your situation, other than shared matches with distant matches like Conor are trickier than shared matches with match like Joseph.
How does confirmation bias factor in when working with mtDNA? I theorize that, in my mtDNA line, a 5th great grandmother’s mother was (let’s call her) Mary Brown. After developing this theory, I revisited my mtDNA test taken years ago. The match list included an exact match with someone whose 6th great grandmother was this same Mary Brown, descending from a confirmed daughter. Since then, another person descending from my 5th great grandmother through a different daughter has tested and is an exact match to me and also the other person confirmed as descending from Mary Brown.
Since working with mtDNA narrows the focus of possible DNA contributors compared to working with autosomal DNA, does confirmation bias factor in when working with mtDNA? How do I tell if I am seeing what I want to see or if I am on the right track in my theory?
Excellent question, Judy. For mtDNA conformation bias isn’t as prevalent in that with mtDNA you don’t have multiple ways you can be connected to other people. The mtDNA gives you one choice: your direct maternal line connects to their direct maternal line. Period. So this DNA connection, combined with your genealogy records, I would say DNA is offering its stamp of approval on this relationship. Now, keep in mind, that the mtDNA match is not good at telling you when you are related. So yes, it could be Mary, but it could be her sister, or her cousin, etc.
This isn’t about confirmation bias. My kids are full blood siblings but one of them has dna matches that the other doesn’t. How is that possible? These matches are on their fathers side, not mine. They share all of my extended familial matches.
Siblings should share all of the same second cousins, and most of the same third cousins, but will have many different 4th cousins. Remember that these full siblings only share half of their DNA with each other. So the other half will find new and different DNA matches.
If a DNA test reveals some one sharing 953 cm’s and 14% same DNA what are the statistical chances of her NOT being a 1st cousin?
There aren’t really statistics for that kind of question. But there are other relationships that would share that much DNA. Like an uncle/niece kind of relationship. There are also other explanations like multiple relationships. For example, if someone is your double first cousin once removed (so two brothers marry two sisters, their kids are first cousins) you might share that much DNA.
Newish to DNA research in my family so a general question. My Irish mother has roots is several parts of Ireland. I have documented reasonably securely up to x3 gens. However since her family left eg Galway (known for its endogamy) over 4 gens ago DNA matches are low (mainly less than 7 but lots around 5cms) so I’m not sure how reliable it is. I also get a lot of high combined cm matches but with multiple small segments. What can I say (generally about this?)
Great question.
If you are trying to connect with people who share your 2X great grandparents, you should be able to find some 3rd cousins. 3rd cousins generally share at least 50 cMs of DNA. If you are trying to connect with other descendants of your 3X greats, those would be 4th cousins and share, on average, 35 cM. But certainly you will have 4th cousins that you share very little DNA with, or none at all.
I just don’t have confidence in matches sharing under 10 cM. You are likely sharing with them just becuase you are all Irish, not becuase you share a recent common ancestor.
You may want to try our DNA and Endogamy course to help you better understand how to look at your match list. https://www.yourdnaguide.com/endogamy
My son and I did the Ancestry DNA test. We both have a match that shows up as my 1st cousin. We both also match his children. However no one on either side of my family that I match and grew up with match that person. How can that be?
Thank you,
Tammy
This is something I would want to take a look at. So you are saying when you click on the Shared Matches you have with this first cousin, you don’t recognize anyone on that list? We could for sure figure it out in a mentoring session, if you wanted to go that route. http://www.yourDNAguide.com/thementor.
I have a friend (or maybe a cousin). We have both tested on Ancestry.com and 23andMe. We do not come up as a DNA match on either site. However we both share about 30 of the same DNA matches. What we know about each other’s history is that going back 100+ years we had ancestors from the same small village in Albania. It appears also that all of those common matches we share are also connected to that same village.
Yes, it sounds like you have endogamy, where you have matches in common because of a common ancestral location. You could be as closely related as 4th cousins, as you only share DNA with about half of your fourth cousins.
My great grandmother was Cherokee Indian and my grandmother and mother all had features of their Indian blood. I have had a DNA test and it does not show any it indian connections at all. It shows that I have 2% from Africa. I have a cousin who doesn’t think I belong to the family. It is very frustrating to hear his comments. I talked with an ancestry rep and he said they are working on Indian heritage but do not have enough DNA at this time. What do you thiink.
Hi Peggy, that can happen sometimes with populations that haven’t been well-represented in DNA testing databases. Our free DNA and Ethnicity guide explains more about why this happens.
I have 2 known DNA matches on 23andMe that share 1.16% (85cM) and 1.11% (83cM) DNA with me. They are siblings and are my 2C1R on my genealogy tree. This fits well on DNA Painter. They have a 1st cousin that is also my match on Ancestry at 45cM (.58%) and he is also a 2C1R on my genealogy tree. On DNA painter he is at 20%, a lower % of 2C1R then those matches that are the same relationship at a higher %, and higher cM. Can you explain this? No one can state that he is not their 1st cousin or my 2C1R.
It could just be that this cousin inherited slightly less DNA from your shared ancestor. Since DNA is randomly inherited, that happens sometimes. Now if you find that you are consistently sharing DNA at a lower rate than expected, then that could be a sign that you are related in a different way than you have been looking at, so just keep that in mind!