which company is right, which is wrong?

[JerryS.]

the same sample read by two different companies, yielded two different results, from the same sample mind you. which one is right, which one is wrong?

Ancestry:
63% Great Britain
5% Scandinavian

FTDNA:
British Isles 4%
Scandinavia 38%

[NCroots]

I believe in essence they are saying the same thing. The British Isles were settled by the Vikings, so it makes sense that Scandinavian would show up. The fact that one shows more Scandinavian and less Great Britain, and vice versa is not really important. It would be important if it was say, Scandinavia and Greek, etc. There is a numerical disparity between the two companies in total, though.

[JerryS.]

I believe in essence they are saying the same thing. The British Isles were settled by the Vikings, so it makes sense that Scandinavian would show up. The fact that one shows more Scandinavian and less Great Britain, and vice versa is not really important. It would be important if it was say, Scandinavia and Greek, etc. There is a numerical disparity between the two companies in total, though.

even if we accept that Scandinavian and British Isles are the same, why such a large discrepancy in the total, 68% vs. 42%?

[keigh]

The DNA that you had tested was the same, but the tests run on that DNA were different. They use different lab samples and computer programs to evaluate and determine the ethnic estimates.

My Ancestry test gives me as 100% European with 57% Great Britain, Iberian Pennisula 12%, Scandinavia 11%, Ireland 10%, Finland and Northwest Russia 7% and a 3% as trace amount of West European. FTDNA gives me as 99% European with a breakdown of British Isles--0%, East Europe--5%, Finland < 1%,Scandinavia--32%, Southeast Europe--0%
Iberia--0%, West and Central Europe--62%

Basically what I know from the tests is that I'm European, and perhaps with Ancestry I know where my ancestors came from, and with FTDNA I know where my ancestors' ancestors came from.

Take the tests with a grain of salt if you have high blood pressure, and a block of salt if you BP's OK.

You might want to download your DNA over at GEDmatch and play around with their ethnicity programs. You may end up even more confused by the result.

[JerryS.]

The DNA that you had tested was the same, but the tests run on that DNA were different. They use different lab samples and computer programs to evaluate and determine the ethnic estimates.

My Ancestry test gives me as 100% European with 57% Great Britain, Iberian Pennisula 12%, Scandinavia 11%, Ireland 10%, Finland and Northwest Russia 7% and a 3% as trace amount of West European. FTDNA gives me as 99% European with a breakdown of British Isles--0%, East Europe--5%, Finland < 1%,Scandinavia--32%, Southeast Europe--0%
Iberia--0%, West and Central Europe--62%

Basically what I know from the tests is that I'm European, and perhaps with Ancestry I know where my ancestors came from, and with FTDNA I know where my ancestors' ancestors came from.

Take the tests with a grain of salt if you have high blood pressure, and a block of salt if you BP's OK.

You might want to download your DNA over at GEDmatch and play around with their ethnicity programs. You may end up even more confused by the result.

using different population groups can reasonably explain why one would test as 20% ABC with one company and 17% ABC with another, but looking just at your results, one company clearly uses suspect calculating data.

[JerryS.]

anyone have a viable explanation for the same data being read/interpreted considerably different ways?

[Boudicca1]

the same sample read by two different companies, yielded two different results, from the same sample mind you. which one is right, which one is wrong?

Ancestry:
63% Great Britain
5% Scandinavian

FTDNA:
British Isles 4%
Scandinavia 38%

Which one most closely matches your known ancestry?

[JerryS.]

Which one most closely matches your known ancestry?

I don't understand. are you saying I should keep testing until I see what I like, instead of simply getting accurate data?

[Boudicca1]

I don't understand. are you saying I should keep testing until I see what I like, instead of simply getting accurate data?

No, I'm saying that sometimes it's helpful for other people to know your ancestry to give you advice based on your original question. I was just trying to help you reach the answer to your question. It's your choice how you pursue finding answers to your results.

[John McCoy]

At the moment, and for the foreseeable future, there doesn't seem to be a rigorous way to validate admixture results. In other words, there doesn't seem to be an external standard by which we could determine which vendor's ethnic origins results are most accurate. Ancestry has reported in a general way on measures they took to validate their methodology, I don't know if any of the other vendors or programmers have released comparable information. But the available tests deal with consistency (i.e., how often do children show ethnic origins that are inconsistent with those of their parents?), which, while an essential attribute of a sound algorithm, is not the same thing as accuracy.

[JerryS.]

No, I'm saying that sometimes it's helpful for other people to know your ancestry to give you advice based on your original question. I was just trying to help you reach the answer to your question. It's your choice how you pursue finding answers to your results.

for now I will say neither, in an attempt to not influence the validity of the data.

[MMaddi]

for now I will say neither, in an attempt to not influence the validity of the data.

Boudicca1's point is a valid one. You ask us which company is right. How can we judge that if you don't give us some idea of your known ancestry to compare with what the two companies have given you in their ethnic/geographic ancestry estimate?

[JerryS.]

Boudicca1's point is a valid one. You ask us which company is right. How can we judge that if you don't give us some idea of your known ancestry to compare with what the two companies have given you in their ethnic/geographic ancestry estimate?

because if I'm 25% German, and 50% English and 25% Italian...., I shouldn't have to say that, it should show within reasonable accuracy if the testing samples aren't skewed or slanted one way or the other. even if the numbers were (based on above) 20% German/West European, 60% British Isles, 20% Italian/Greek/Southern European.... the previous poster (http://forums.familytreedna.com/showthread.php?t=41886) had an Ancestry report of 1% Caucuses (West Asian listed on some modules) but almost 10% on GEDmatch. there is a huge difference between barely showing (1%) and clearly there (10%) vs. 30%-40%.

[Germanica]

What it comes down to is none of them are "right" or "wrong". They are ALL just estimates, and depending on your particular DNA, what particular sample groups the company happens to have (and that doesn't mean there's a deliberate slant in any direction, it's just a matter of how your DNA matches up with their sample groups), and how they process/analyze the data can mean that one company/calculator matches the known ancestry of one person better, while a different company/calculator is more accurate for another person. There is no one "right" company/calculator for everyone. That is why people need to know your ancestry to say which one matches yours better. The problem is you're expecting a level of accuracy and precision from an imprecise science. The accuracy you're looking for just doesn't exist with ethnicity reports, sorry.

[ada]

because if I'm 25% German, and 50% English and 25% Italian.

Well that's certainly curious. I am an immigrant from Poland, very Polish on all sides. Yet when I uploaded my Ancestry's raw DNA to FTDNA I got the following result in My Origins:

European 96%
- East Europe 90%
- British Isles 6%

In reality, my liberal estimate is that maybe 6% of my ancestors knew what Great Britain was. In other words, it's complete baloney. Yet you, who have a British parent, are apparently 2 percent less British than me.
Ancestry DNA correctly estimates that I am 0% British and it makes sense.