Being off by a few percent from nothing more than a cheek swab doesn't seem like a sham. I'm sure if they got a lot more sample to test from it could be made more accurate.
Have you looked at things like thermometers? They don't offer more than +- 2F accuracy for those under $100. Some meters, etc only offer +- 5% accuracy or worse. Those aren't shams.
In the article, one service reports 61% Balkan and another reports 23%. That one algorithm is mostly accurate within itself is hardly notable. That they're not close with one another means that one of them is way off or the industry is very inaccurate. If the latter is correct yet they're purported to be reasonably accurate, it's a sham.
Some of the difference might be classification differences. Balkan vs. Eastern European vs. Russian isn't hard and fast.
My wife's percentages have changed over time w/o a new sample- they've just refined their data and can give better results from the original data.
Sure, but does it account for a large percentage change? If their reference data is so limited that adding to it can result in monumental changes, then it is currently inaccurate and a sham. If it isn't a large percentage change then lack of refinement doesn't adequately explain the significant variance across companies.
That reminds me I read somewhere that ovens in residential kitchens are within 20 degrees F of the stated temperature, either because of the position in the duty cycle for the heating element or noise or bias in the temperature measurement. But no one decries ovens as shams.
What you give up (your dna) and what you get back (ancestry data of a dubious validity) doesn’t seem worth it to me.
That doesn't make it a sham.
Yes it does. They are selling it as DNA = Ancestry. For me, that is blatantly misleading at best and fraudulent at worst.
In my opinion, yes. It's a relatively cheap entertainment genetic game. For me, it's akin to playing the lottery or getting an astrologer to read your aura.