Why do ethnic tests often fail?

 People often wonder why different genetic tests give them completely different ethnic profiles. Finns sometimes get either a completely or almost completely Finnish profile, sometimes, for example, a profile that is only 60% Finnish and 40% northwestern European, such as Scandinavia and Britain. This is where many people's faith in these tests falters. Then the know-it-alls raise their own profile and competence by defending the results with their ideas. However, it is clear that the same person cannot get such different results as we see. The thing is that the results depend solely on the method used. Which method the test developer has used depends solely on the intended purpose of the test and the preferences of the developer. I will now go through the test methods categorically, not the purposes, because few test developers reveal them. The methods are mainly divided into three groups:

1 Tests based on allele distribution. In these, a Finnish test taker often gets a very wide geographical distribution, which can vary depending on the test references. For example, I can get 60% Karelia + 40% Germany or 60% Estonia and 40% Sweden or an additional 5% Greece or Southern Italy.

2 Tests based on allele distribution profiles. In these, a Finnish test taker can get up to 100% Finnish results, but only when the test distribution exactly matches the allele distribution of the ethnic samples used in the reference. Usually, the results of Finns include some minority ethnicity, which is typically 10-20%.

3 Tests based on allele or haplotype distribution weights. In this way, you can almost always achieve 100% ethnic identification, but as in point 2, only for part of the ethnic profile. The allele or haplotype distribution can be optimized to maximize ethnicity readings. On the other hand, weighting the data can also lead to incorrect results, where some minority ethnicity is dropped from the results. In these tests, I usually get a result of 95% Finland + 5% Sweden.