Ethnicity Testing- Is it really worth it?

[Bob Spiers]

I've searched new posts but can't find your post Bob. Perhaps admin don't want WW3 to break out!

"Awaiting Moderation" is the keyword with General Genealogical Queries postings which often take a while. A bit like Private Tree messaging (sorry as I know you are a fan), one has to wait for someone to be at home.

 

[FamilyHistoryGal]

If you message someone with a public tree to query an error or for further info you still have to wait for a reply (Tee-Hee)! Give me a heads up when the thread is active

 

[peter]

Haven't we already had this discussion, Bob - or did I imagine it?

 

[Bob Spiers]

Haven't we already had this discussion, Bob - or did I imagine it?

No, and I have said as much with a comment about you and I agreeing to disagree. But as you will see if you go back a page, at least two other members think it time to air it again, and I am sure others will be of the same opinion as it is a subject that interests many and has advocates on both sides of the argument. I have therefore (as mentioned to FamilyHistoryGal) opened the debate under the heading: "Ancestry Public Trees versus Private - a new Debate" in favour of Public Trees , which at the moment is hidden as it awaits Moderation, as you will know.

Once released I am sure it will (again) make for good debate with support for both sides of the argument as all good debates do.

 

[Bob Spiers]

Give me a heads up when the thread is active

Yes now up and running FHG as Ancestry Public Trees versus Private - a new debate. Look forward to reading your response, and I expect Jorghes also. Can't wait.

 

[Bryman]

someone I worked with back in the 70's & 80's by the name of Lionel Christmas

Thanks Bob. Sorry this might appear a little late but Spring has arrived here in NZ and everything is growing so lots to do to keep on top. Now the forecast is for a wintery blast up from the South over next couple of days so double workload. Thankfully not like Michael currently approaching Florida.

I have looked at records for Lionel and family but as with rest of that surname I find too many individuals to definitely isolate any that are relevant. I will get back to you privately if there is anything else to say/discuss rather than take up further space off-topic, unless I get a reply from another member to make a discussion split worthwhile. Thanks for tip-off.

 

[Helen7]

Unfortunately, those with a match to me don't always show a match to either of my parents in the Ancestry comparisons.

Just picked up on this post from last year which seems to answer a question I posed on another forum about whether it's worth getting my sons to test. I assumed they would just get a combination of my matches and my husband's, with no extra ones appearing, but your experience jorghes suggests otherwise. Can someone with more knowledge of the DNA matching system than me tell me why this should be so? If children get ALL their DNA from their parents (as I believe is the case), any extra matches would seem illogical, unless the extra matches are spurious?

 

[peter]

If children get ALL their DNA from their parents (as I believe is the case), any extra matches would seem illogical, unless the extra matches are spurious?

Generally you'd be right, but bearing in mind the whole process is about balancing false positives against false negatives there will be times when a genuine match which was incorrectly classified when the parent was tested could flip, simply by chance.

Something else to consider is that Ancestry use a phasing algorithm which, whilst supposedly highly accurate, can't be perfect. This could lead to genuine matches being missed at the parent level, but picked up at the child level. (I can explain why, but I'm not going to as it's rather complex and I have a newsletter to write!)

And finally, whilst atDNA tests do produce very accurate results - as you can see if you test with more than one company - there will be a small number of errors and missing bases (see this blog post). If one of those just happens to be in the middle of a shared segments.

But is it worth the money - almost certainly not. You would do better to spend the money testing siblings or cousins (of whatever order).

 

[Helen7]

And finally, whilst atDNA tests do produce very accurate results - as you can see if you test with more than one company - there will be a small number of errors and missing bases (see this blog post).

Thanks for this. I read the blog, and as I've uploaded raw DNA results to GEDmatch I completed their survey.

 

[canadianbeth]

I completed the survey as well. It seems that my "no-call" numbers were quite low.

 

[peter]

My no-call figure was 0.48% for my Ancestry kit, but half that for my FTDNA kit. I shall have to upload my 23andMe results to see how they compare.

 

[canadianbeth]

Mine was 0.116% for Ancestry. I just uploaded to FTDNA today. Can I upload to 23 and me as well? I looked at the site but did not see where it could be done.

 

[Helen7]

Mine was 0.06% and my husband's 0.11% (both Ancestry kits) which I though was quite impressive. The no. of matches also seemed large (16,490 >= 7 cM for me and 13,774 >= 7cM for him, though still not as many as with Ancestry)

 

[peter]

Can I upload to 23 and me as well?

No - supposedly there was a short period when 23andMe accepted Ancestry results but it was long ago and I only found out about it last month. I tested with 23andMe when they had a special offer in the UK earlier this year.

In theory I ought to be able to take my three sets of DNA and produce a set of results with even fewer no-calls. But unless there's a program that'll do it for me, I don't suppose it will ever happen.

 

[peter]

The no. of matches also seemed large (16,490 >= 7 cM for me and 13,774 >= 7cM for him, though still not as many as with Ancestry)

I'm not sure how meaningful those statistics are - according to GEDmatch my FTDNA kit has 21923 >= 7cM and 55451 < 7cM but my Ancestry kit has 12289 >= 7cM and 34926 < 7cM

Considering that GEDmatch only has about 1 million users both sets of figures seem way too high.

 

[Helen7]

Considering that GEDmatch only has about 1 million users both sets of figures seem way too high.

I agree, especially as the matches that I've found in common in both Ancestry and GEDmatch always have a higher cM count in GEDmatch - presumably because of Ancestry's algorithm filtering out some false positives. I presume if I upload to MyHeritage I'll get different results again?

 

[peter]

I presume if I upload to MyHeritage I'll get different results again?

Some of the people in the MyHeritage database - perhaps the majority of them - will have tested with other providers and uploaded their results. I certainly recognised some of the names amongst my closest matches.

So far I haven't made any groundbreaking discoveries, but I've yet to contact of my matches in Germany.

 

[palfamily]

I found an Argentinian cousin on MyHeritage, which I was hoping for when uploaded my DNA results.

 

[DavidL]

MyHeritage results are back. I seem to have developed a Scandinavian past, making up 27% of my ancestry. Not a smidgeon of Vikinghood on Ancestry, LivingDNA nor FTDMA. Perhaps it's why my new grand-daughter is a blue-eyed blond in a family of dark-haired, dark-eyed relatives. Or perhaps it is unwise to give much credence to ethnic tests. On the other hand, they are all consistent about my Ashkenazi roots.

 

[jorghes]

MyHeritage results are back. I seem to have developed a Scandinavian past, making up 27% of my ancestry. Not a smidgeon of Vikinghood on Ancestry, LivingDNA nor FTDMA. Perhaps it's why my new grand-daughter is a blue-eyed blond in a family of dark-haired, dark-eyed relatives. Or perhaps it is unwise to give much credence to ethnic tests. On the other hand, they are all consistent about my Ashkenazi roots.

Snap.

I also have suddenly gained a Scandinavian Ancestry of 39.5%, topping my English (26.3%) and Irish, Scottish and Welsh (23.8%); they did get my Ashkenazi Jewish branch (9.6%)... but I am also (apparently) 0.8% South American Amazonian... I am unsure where the Scandinavian dominance has come from and while I do have cousins who did live in South America, none of them would have contributed to my DNA.

The most interesting result since LivingDNA attempted to use a dodgy sample and told me I was from Africa. (LivingDNA doesn't pick up my Jewish branch, but then I didn't test with them to find out more about the Jewish side.)