- Posts: 1
After having DNA evaluation done by 2 of the most marketed test providers, I was disappointed by the conclusion differences. The big percentages were in the ballpark... but the differences in the 1% to 5% geographic and ethnic result conclusions were completely different.
Not easy to find the discussion on this but guessing the conclusions are driven by each organizations database. It is a correlation conclusion that is being made in many of these cases... not some objective dna tag that conclusively connects to a specific ethnicity. (Possibly with some exceptions for Ashkenazi or other group that has stayed pretty isolated in their unique community.)
Can anyone direct me to resources or research that addresses this?
- Posts: 850
Family Tree: Graphic
View their family tree
The main issue is that we're using a sample base of contemporary testers to estimate ancestry thousands of years ago. Differences in the sample base, the self-identification of the testers, and the algorithms used explain why each company comes with different results. At a continental level, it works fairly well. At a more regional level, there has been so many migrations and the DNA is so mixed that it's hard to distinguish English from German, etc.
This article goes into a bit more details https://www.livescience.com/62690-how-dna-ancestry-23andme-tests-work.html