4 ms·
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 mo
by docker_up 8y ago
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.
- kodablah 8y agoIn 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.
- mrkstu 8y agoMy 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.
- kodablah 8y agoSure, 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.
- jnwatson 8y agoSome of the difference might be classification differences. Balkan vs. Eastern European vs. Russian isn't hard and fast.
- glup 8y agoThat 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.