It ranks right up there with an electronic version of the OUIJA Board.
SA is now a Gorebal Warning fish wrap.
My purported ancestry initially included traces 1-3% “North African” and “European Jewish”
Recently the “new and improved” data eliminated both of those.
Just sayin
How Accurate is Scientific American?
If you can find distant cousins and the tree to prove it you have much more science than the total science behind global warming.
the gold standard is Nature and other journals - SA is a political but scientific global climate change joke right up there with Nat Geo.
I think it is “Scientific American,” not “Scientific America”...(?)
Ancestry DNA tests are fudged and not at all accurate or truthful.
I posted this yesterday:
Here’s a fun fact yet to be brought up: These “results” will change over time even on a specific DNA ID site. ancestry.com specifically makes a big deal about how a specific person’s results from some number of months ago get “updated” and presumably more accurate because of more samples and customers since then.
I am looking at a particular set of “results” there at the moment. I see that five sources of DNA, one of those sources a full seven percent, are “no longer in estimate,” which was made roughly seven months ago.
So any talk of ancestry of 1/64 (1.5%) being statistically significant, far above any margin of error, unimpeachable, proof positive of ancestry, etc., is simply mumbo jumbo.
I read an article somewhere not that long ago saying the folks running these $DNA$ find-your-roots companies (they do some nice advertisements with people traveling to exotic places to visit their ‘ancestors’) throw in various results that are not related to the actual DNA. Those paying money for results want results that are “interesting” and they’ll give you that.
Despite all the hoopla, such as identifying a killer in California, the science of DNA data mining is in its infancy. Sloppy procedures, misreading, and the like aside, 20 years from now, the information gleaned will be fantastic.
They are pretty accurate in my eyes - they detected Elizabeth Warren was 100% pure idiot.
For European ancestry, I’ve found it to be fairly accurate based on my family’s genealogical research. My siblings and I used ancestry.com’s databases, which recently updated and refined their results due to their reference populations growing from ~3,000 samples to ~16,000.
According to them, I’m 43% Irish/Scot, 37% French, 14% Spanish and 6% Italian, which is just about exactly correct based on my immigrant background.
Yes, I’m a European mutt... :-)
My good friend from West Virginia has a maternal grandmother who was Cherokee. But the DNA test said that my friend had no Native blood. That could only be true if either she or her mother were adopted.
My grandfather (also from WV) was 1/2 Cherokee, but I’ve no desire to have my DNA tested. I believe something nefarious will come of it at some point.
All of these tests are limited in what they can find. The Y-Chromosome and Mitochondrial DNA tests are good and they may be good at finding very close relatives and identifying lost ones, but the percentages of the nationalities of ancestors is just a rough guess.
They seem to do a decent job on a continental level. They can find the presence of an African ancestor in a European or a European ancestor in an East Asian. But separating Norwegian DNA from a Spaniard is more difficult and separating Irish DNA from an Englishman is getting extremely hard.
My dna sample results from ancestry match our family’s research and lore.
“Scientific America is a peer review journal that some hold up to be the gold standard.”
B.S. I subscribed to Scientific American for over two decades, and kept every edition in a large magazine collection, when it was then a good science magazine - until sometime in the 1990s when the “scientific” level of its front cover articles became no better than the History Channel’s “Ancient Aliens” cable TV show. I quit the subscription, and checked new editions at the magazine stand, remaining convinced I was right to cancel the subscription. After three or four years of that I quit checking it out. I would not rank it today as a “bronze” level science journal, much less “gold”. I don’t know who told you it was a “peer reviewed” journal. It isn’t.
This is a very good article that helps explain a few basics. First the DNA tests offered by Ancestry and 23 and Me provide you with an “ethnicity estimate”. This is marketing pap. The results are a function of their own databases and how they are applying probability to certain data. If you take both tests, you will get different results. I have gotten two very different results from one Ancestry sample as they have enlarged their own database and changed their methodology. Not serious science.
Two, I am very skeptical using these tests for medical analysis as is done by 23 and me. It’s another marketing trick, but one that could have detrimental downsides depending how the information is used.
Three, these tests are useful for genealogical purposes, especially among populations that share a common surname, lived in the same area, and family lore has confused who is whom. To do this, you need to share and compare results with other potential cousins, and you usually need to use third party software tools that will allow you to do the analysis. Ancestry, for example, doesn’t let you know how they are making the stew, you just have to trust their result. One downside is that other people get a copy of your results. It’s not your entire genome, but it’s the interesting bits and there is some risk. Since the DoD already has my DNA, this isn’t a big risk from my perspective, but others might have a different thought.
All I know is, Ancestry matched me up with relatives on both my mother’s and father’s side, and even knew the correct relationships. How would it know, unless there’s something to this?