#1
|
|||
|
|||
Ethnicity estimates
What on earth?
OH's DNA - 79% Scottish - ( NE Scotland & Northern Isles, Fife & Angus), Ireland 18% and 3% Wales. There should be lowlands Scotland and Northern England through the places his family were born c1800 Edinburgh/Midlothian and Berwick on Tweed but not a mention. Son's DNA is 58% Scottish (NE Scotland & Northern Isles, Fife & Angus) 40% Irish and 2% Norway However my son's ethnicity estimate gives no English or lowland Scottish heritage at all. I can understand the Irish bit through my father but there's nothing at all of my mother's Lancashire/Yorkshire heritage (only her grandmother was born in Ireland in 1860 the rest born in Lancashire or Yorkshire) and match up with other DNA matches going back to 1800. Confused to say the least.
__________________
|
#2
|
||||
|
||||
Think of all those Ancestry trees and their accuracy. Ethnicity is only a toy. Choose another provider, with a different database, and the results will be quite different.
If you drill down into the results, they give a range of certainty. I allegedly have 10% Scottish (not supported by the paper trail back to 1800) But the range is 0 - 29% My marginal ethnicities have fluctuated wildly from year to year: Spanish, French, Swedish and now Welsh. Hopefully it proves more accurate over time, but it does depend on the accuracy of underlying research.
__________________
The chestnuts cast their flambeaux |
#3
|
|||
|
|||
Just have to tell my son that according to ancestry he has no English roots at all. He'll be happy with the Scottish results though lol.
__________________
|
#4
|
||||
|
||||
All you can say at the moment is that the majority of his heritage is Scottish/Irish, which agrees with what you know.
__________________
The chestnuts cast their flambeaux |
#5
|
|||
|
|||
Yes but I still think there should be around 25% English but there not being even 1% in the estimates means its just guesswork on their part.
__________________
|
#6
|
||||
|
||||
I don't think it works like that, Julie. In theory, we could inherit virtually no DNA from one grandparent, and nearly 50% from the other. I have certainly appear to have got much less DNA from my mother's mother than I have from her father.
__________________
The chestnuts cast their flambeaux |
#7
|
||||
|
||||
He still has the English heritage but he can not possess all the DNA from all his ancestors.
Also if your relatives came from Ireland say 500 years ago, when many of us can't get back 200 years, the DNA might say Ireland, even though you believe there they always lived in one tiny English town.
__________________
Toni |
#8
|
||||
|
||||
My percentages change every time they update it. I now have 23% Scottish (which was hardly a thing when I got my first analysis) and the 3% Swedish, which disappeared, has now turned into Swedish/Danish 11%.
I can't account for any of that but the 59% English, broken into Cornwall, Gloucestershire central southern England and East Anglia does fit with my research.
__________________
Love from Nell researching Chowns in Buckinghamshire & Oxfordshire Brewer, Broad, Eplett & Pope in Cornwall Smoothy & Willsher/Wiltshire in Essex & Surrey Emms, Mealing + variants, Purvey & Williams in Gloucestershire Barnes, Dunt, Gray, Massingham, Saul/Seals/Sales in Norfolk Matthews & Nash in Warwickshire |
#9
|
||||
|
||||
I have come to the conclusion that my family do not have any DNA in common with some of those on the reference panel.
Therefore, our ethnicity estimates will not be in any way accurate until the Reference Panel changes to include some of those we share DNA with. I have no idea how it is concluded that my Dad has Scottish ethnicity, as none of his lines lead anywhere near Scotland. Do they just guess or extrapolate if there is no common DNA to cover a certain percentage? This page explains how the Reference Panel is comprised of 56,5800 people, not a very large sample, I would have thought - https://www.ancestry.com.au/cs/dna-h...city/estimates |
#10
|
|||
|
|||
It appears that many people are displaying Scottish ethnicity when none is proven by research. To my mind, this flags up a problem.
|
|
|