4 ms·

1. That’s not statistics, it’s just a ratio. 2. What matters is whether or not we can reproduce with chimpanzees. We can’t, so we’re not chimpanzees.

by dqpb 5y ago

brudgers 5y ago
The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. In terms of reproduction, my understanding is that modern populations suggest that our recent ancestors could mate with Neanderthals. And did. Of course reproducing is a messy criterion beyond that. Children and old women and just some regular folk can’t reproduce but are people nevertheless. And in vitro and surrogacy are another can of worms.
- Dylan16807 5y ago
  > The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. Confidence based on what math? Species isn't based on raw percentage of matching DNA. Nor is that how p values work. > In the extreme it is 23andme emphasis on “18% Welsh but I thought I was English” as scientific. Every report ought come back as 98.8% Chimpanzee if it was science. That calculation is comparing only the genes that differ between the references, and weighting them across 100% to give an easy to understand result. If you added a chimp into the reference pool, they'd show up as less than 1% every time. If you added a chimp into the reference pool, and removed all the human samples except welsh, then the test would say "on a scale from chimp to welsh, you are >99% welsh". But that's a matter of lacking the data to be more precise, not because the method is fatally flawed. The place where flaws might come in is when the source data is mis-labeled, not because the math is wrong. > Sure there’s a little science. But mostly it is theology. A belief that humans are especially special masquerading as science. No it's not. It would do the same thing with chimps. Put a bunch of them in the 23andme database and then a sample chimp will come back as 25% this area chimp, 40% this area chimp, 34% this area chimp, 1% match to humans.
  - brudgers 5y ago
    Welshness is entirely cultural not at all genetic.
    - Dylan16807 5y ago
      There's definitely a correlation between welshness and specific groups of genes. Maybe "welsh" is too fine of a distinction to determine. But you can easily tell the difference between that general area vs. native american vs. subsaharan african ancestry.
      - brudgers 5y ago
        Someone can be Welsh by adoption or migration. Even if they were born in Belize, South Africa, or Trivandrum.
        Dylan16807 5y ago
        Sure if you want to ignore what the test is for. But uh my main point was to explain how the issues have nothing to do with distinguishing chimp from human, and there is no reasonable metric that would say they're the same species, and I wish you wouldn't ignore my main point to talk instead about the definition of welsh.
- dqpb 5y ago
  > The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. Do you really think this or are you trolling statisticians?

1. That’s not statistics, it’s just a ratio. 2. What matters is whether or not we can reproduce with chimpanzees. We can’t, so we’re not chimpanzees.

by dqpb 5y ago

1. That’s not statistics, it’s just a ratio. 2. What matters is whether or not we can reproduce with chimpanzees. We can’t, so we’re not chimpanzees.

brudgers 5y ago

The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. In terms of reproduction, my understanding is that modern populations suggest that our recent ancestors could mate with Neanderthals. And did. Of course reproducing is a messy criterion beyond that. Children and old women and just some regular folk can’t reproduce but are people nevertheless. And in vitro and surrogacy are another can of worms.

Dylan16807 5y ago
> The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. Confidence based on what math? Species isn't based on raw percentage of matching DNA. Nor is that how p values work. > In the extreme it is 23andme emphasis on “18% Welsh but I thought I was English” as scientific. Every report ought come back as 98.8% Chimpanzee if it was science. That calculation is comparing only the genes that differ between the references, and weighting them across 100% to give an easy to understand result. If you added a chimp into the reference pool, they'd show up as less than 1% every time. If you added a chimp into the reference pool, and removed all the human samples except welsh, then the test would say "on a scale from chimp to welsh, you are >99% welsh". But that's a matter of lacking the data to be more precise, not because the method is fatally flawed. The place where flaws might come in is when the source data is mis-labeled, not because the math is wrong. > Sure there’s a little science. But mostly it is theology. A belief that humans are especially special masquerading as science. No it's not. It would do the same thing with chimps. Put a bunch of them in the 23andme database and then a sample chimp will come back as 25% this area chimp, 40% this area chimp, 34% this area chimp, 1% match to humans.
- brudgers 5y ago
  Welshness is entirely cultural not at all genetic.
  - Dylan16807 5y ago
    There's definitely a correlation between welshness and specific groups of genes. Maybe "welsh" is too fine of a distinction to determine. But you can easily tell the difference between that general area vs. native american vs. subsaharan african ancestry.
    - brudgers 5y ago
      Someone can be Welsh by adoption or migration. Even if they were born in Belize, South Africa, or Trivandrum.
      - Dylan16807 5y ago
        Sure if you want to ignore what the test is for. But uh my main point was to explain how the issues have nothing to do with distinguishing chimp from human, and there is no reasonable metric that would say they're the same species, and I wish you wouldn't ignore my main point to talk instead about the definition of welsh.
dqpb 5y ago
> The null hypothesis is chimpanzees and humans are not the same. We can reject it on the basis of DNA with greater than 95% confidence. Do you really think this or are you trolling statisticians?