However, we have out own version of the chicken and eggs dillema don't we?
It is an unfortunate necessity to have to rely on the historical and ancestral record in order to develop the technology in the first place. For example you have to test a statistically significant population of subjects with a "known" ancestry to begin with in order to develop the SNP patterns that relate the DNA to it's geopgraphical origin. Once that is done it is necessary to validate the results by testing new subjects and comparing the predicted results to their "known" ancestry.
At some point you strive for a degree of confidence in your results that allows you to tell a subject that his own understanding of his ancestry is flawed.
My question is, how is that determined? Since it is necessary to rely on the "known" ancestry that by it's very nature does not agree with the DNA results.
News 
Market Data 
Discover 
