Genealogy and family history are hobbies and passions that have been around for millennia. However, recent technologies like DNA testing and online record databases have reinvented the search for knowledge of one’s heritage and have made more options available to the general public. The company 23andMe is best known for its DNA ancestry kits that connect you to distant relatives and display the percentage of the different ethnicities that make up your ancestry. But just how does this company predict this information and how can we trust that it is accurate? 

How biomedical ancestry services work: 

23andMe uses modern biological technologies to determine ancestry composition, genetic relatives, and health inclinations for a given person’s sample. When you send in your DNA testing kit, 23andMe acquires your DNA traces through your saliva. The DNA composition is deconstructed to locate individual segments of DNA called genomes. These genomes are compared with their reference pool, which is a set of 14,000 individuals’ DNA whose ancestry and medical data is already known. When 23andMe’s algorithm detects a DNA segment in another person that resembles one in yours, it compares them to determine their position on your family tree. This is obviously not perfect, but it renders relationship suggestions that are close to reality, within practical limits. Your final overall ethnic composition is a result of the system adding up individual comparisons with known DNA samples to produce an overall estimation of your ancestral geographic make-up. 

How the information could be misleading: 

Although this company manages to find incredibly useful biological information, it is important to note that their geographic data is not exact. It is theorized that modern humans have been wandering around the Earth for about 70,000 years, when homo sapiens first emerged as the dominant hominid species and migrated out of East Africa. It was not until around 10,000 years ago that these groups moved away from their nomadic lifestyle and began to settle down in specific geographic locations. Since then, several factors like famine, natural disaster, and political unrest have pushed people to migrate elsewhere. Therefore, it somewhat problematic to suggest that one person is completely from one geographic location, because it ignores the fact that all humans have migrated at one point or another in history. 

While the ancestry results produced are the best we have for now, they are not perfect. It is important to keep in mind the potential sources of error one might encounter in one’s results. What are the ways in which the results could potentially go wrong? There are two ways detailed on 23andMe’s guide. The first is that there is so much admixture between humans that separating distinct categories of ethnicities is a gargantuan task. Haplogroups (descendants of groups that emerged from different migration route out of Africa about 70,000 years ago) are, for now, the best estimate one can make about categorizing different groups. For example, Latin Americans in South America today are likely to have an admixture of European (Spanish, Portuguese, etc.) and indigenous native (Aymara, Quechua, Guarani, etc.) ancestry. There may also be a lot of addition of slaves and indentured workers from West Africa and the Indian subcontinent that were imported to plantations early in the continent’s colonial history. Even those ethnic groups within smaller geographic regions have been intermingling for millennia. 

The second potential for error arises from the fact that current technology cannot make a distinction between which chromosomes come from which parents. Humans have 23 pairs of chromosomes, with 23 chromosomes inherited from each parent. The way the chromosome pairs are linked, current technological limitations do not have a way of definitively determining which comes from the father and the mother. This makes it difficult to discern which side of the family certain features come from, including physical traits as well as ethnic bio-history. 

How this information contains biases: 

A very commonly heard criticism of 23andMe and several other ancestry websites and companies is that their tests tend to be Eurocentric, fostering racial inequity. Considering this, it becomes more apparent that these DNA tests were in fact designed for Americans of European descent. As the country itself is only a few hundred years old, a great portion of the incoming immigrants arrived in the 19th century. Many white Americans are a mixture of ethnicities only going back about two centuries. This makes it much easier to compare the DNA of current Europeans and white Americans, the main customer base of these products. There are other ways in which this test is heavily Eurocentric. For example, most of the reference populations that 23andMe uses to compare DNA are derived from modern day borders of countries. This is an issue for those of African descent as the borders of this continent were drawn primarily by European powers in 1884 at the Berlin Conference. These boundaries completely disregard the tribal ancestry and culture that still exist in Africa today and categorize groups in rather non-uniform and inconsistent ways. 

As much as this has social implications that cannot be dismissed, it is also important to remember that this is more of a practical issue than social. 23andMe does acknowledge the potential for bias in their data and has pledged to constantly improve their database with more diverse samples as available. The problem is, historically, most biomedical information saved has come primarily from European males.  

There are many other problems with the samples too. For example, the sample size for different categories is most certainly not proportional to their relative population sizes worldwide. There are far too less numbers of Indian and Chinese subjects tested in the reference samples to be statistically relevant given that these two “ethnicities” make up about 36% of the world population (what makes up an ethnicity is debatable). Some major groups with severe social influences are left out of consideration entirely, while other minor groups with much less of a compositional influence are included in the samples.  

Alternatives for researching family history and genealogy: 

Although these tests contain biases and some misleading information, we believe that they are still somewhat useful to those who are interested in learning more about their heritage and biological information. The test can be very pricey though, and there are some alternatives to researching your history further. With a Penn State access account, the Ancestry Library database is available through the PSU Libraries’ website. Although the Ancestry database also gravitates towards Americans and Europeans with most of their records being from the U.S. Census, it may be a useful alternative since it is free to Penn State students and faculty. Another free alternative is the website Family Search. It is a free service where you can create your family tree. When entering a name and birthdate of your ancestor, it attempts to match the information to family trees that may already exist on their site. Aside from these more official websites designed for genealogy, our library has some resources that may also be useful in this process: the library’s genealogy guide, Penn State’s State College Census Project with historical Sanborn maps from 1920 and 1930, and the Eberly Family Special Collections. 

Olivia Neill is a rising Junior Geography major who is also pursuing minors in Spanish, History, and Political Science. Her interests are in human geography, anthropology, and GIS. She has been working at the Donald W. Hamer Center for Maps and Geospatial Information since January 2021.  

Ishaan Anavkar is a rising Senior majoring in History and International Politics. He is also pursuing minors in Geography, Linguistics, and Anthropology, as well as certificates in Human Geography and GIS. He has been working at the Donald W. Hamer Center for Maps and Geospatial Information since January 2019.