3 ms·

This is a great reason why I won't get this test done. The math is bad and (genetically identical) twins getting different results shows how bad the science is

by socrates1998 8y ago

This is a great reason why I won't get this test done. The math is bad and (genetically identical) twins getting different results shows how bad the science is right now. This reminds me of when economics try to predict what the GDP will be next year. It's almost always wrong, so why would ever listen to anyone who is always wrong? It's mostly a scam, in my opinion.

devereaux 8y ago
Actually, the results are not so bad. The technique is rough, but even some basic MDS (multi dimensional scaling) on DNA gives things that look like the actual maps, so there is some scientific base. Also, you should not underestimate how not knowing where you ancestors came from is disturbing for many people. EDIT: for those who disagree, check what just the 1st two PC can give for European DNA after a simple rotation of the axis to match a map : http://blogs.discovermagazine.com/gnxp/files/genmap1.jpg http://blogs.discovermagazine.com/gnxp/files/genmap1.jpg I think it's fair to say there is a scientific base.
- ralusek 8y ago
  Why should that be disturbing to some people? I think the vast majority of people, at least in America, have no idea where their ancestors came from. Is the usual "I'm part Irish part German part Cherokee" meaningfully more satisfying than "I'm west African?" I don't think it's a good thing for people to attempt to derive any aspect of their identity from their genealogical ancestry, it has no bearing on who you are as a person, and has the potential for great harm. I've seen it used to give people a false sense of culture and belonging to a community that they previously had nothing to do with. If you felt a connection to a particular culture or identity, should we really be telling people that their genome is responsible for determining whether or not they can associate parts of those cultures with themselves? Racism results from a failure to separate identity from genome. If you listen to the dogma of a white supremacist, it's heavily predicated on associating the accomplishments of their "European" forefathers with themselves. If you listen to the dogma of a black supremacist, it's heavily predicated on assuming the identity of the Egyptians or Israelites and associating their accomplishments with themselves. I really think that if an individual is truly "disturbed" by not knowing the precise nature of their lineage, that is an individual in search of an identity on a plate, and therefore highly corruptible. The motivation to learn more about one's ancestry should ideally not surpass the level of an intellectual curiosity.
  - devereaux 8y ago
    I wholeheartedly agree with your analysis on the risks. I still think it can have a positive role, if only to help the person emotional investment. > Is the usual "I'm part Irish part German part Cherokee" meaningfully more satisfying than "I'm west African?" For that, "I'm west African" is less satisfying than "I did trace my roots to Mali" which can lead the person to study the history of the country, Timbuktu, etc.
- socrates1998 8y ago
  These tests aren't a measure of ancestors, it's a measure of a person's DNA and where that DNA possibly comes from. Theoretically, they are similar, but since you inherit a random 50% of DNA from both parents, on an individual scale, you might inherit a misleading percent from your ancestors. Let's say 3 of your grandparents are all full-blooded Italian. Like, their families have been in the same village for centuries. And one of your grandparents is from England (and is similarly from an old family in a small village). So, theoretically your DNA results should 75% from that region of Italy and 25% from English region. However, lets say from your mom, the parent who is 50% English and 50% Italian, you only inherit a mix of 15% Italian and 35% English (15% + 35% = 50% mom's DNA). It's random, so this is possible, especially on a small sample of ONE person. And from your Dad, who is 100% Italian, you obviously inherit 50% of your DNA as from that Italian region. Now, you theoretically should be 75% Italian and 25% English, but since it's random, your true DNA reads as only 65% Italian and 35% English. This is why it's mostly bullshit to me. It's impossible to know what percents your DNA randomly passed on to your ancestors. Sure, in our example, the two places where correct, Italy and England, but the percents are way off. And that's honestly WHY we get these tests. We generally know where our families come from, we just have no idea what the percents are. So, the whole point of the service is stupid since they equate DNA with ancestry.
  - devereaux 8y ago
    > We generally know where our families come from No, not everybody. Even percentage that are a bit off are better than nothing at all.
    - 0xffff2 8y ago
      Why though? By ancestry, I'm 25% Italian and a whole lot of other stuff. By percentage, I think I'm probably either mostly French or mostly English by ancestry, but that doesn't matter to me at all. What matters to me is that I grew up learning to cook from my Italian grandmother. As a result, in so far as I have any cultural identity other than "American", that identity is Italian, regardless of my genetic origins. It's hard for me to imagine why anyone would feel any significant connection to an ancestry that they needed a DNA test to know they had.
empath75 8y ago
I sent my grandparents raw dna file to a few different sites and got wildly different results, including 1% nigerian and 10% north african from myheritage.com (something that didn't show up at all on other sites.)

devereaux 8y ago

Actually, the results are not so bad. The technique is rough, but even some basic MDS (multi dimensional scaling) on DNA gives things that look like the actual maps, so there is some scientific base. Also, you should not underestimate how not knowing where you ancestors came from is disturbing for many people. EDIT: for those who disagree, check what just the 1st two PC can give for European DNA after a simple rotation of the axis to match a map : http://blogs.discovermagazine.com/gnxp/files/genmap1.jpg http://blogs.discovermagazine.com/gnxp/files/genmap1.jpg I think it's fair to say there is a scientific base.

ralusek 8y ago
Why should that be disturbing to some people? I think the vast majority of people, at least in America, have no idea where their ancestors came from. Is the usual "I'm part Irish part German part Cherokee" meaningfully more satisfying than "I'm west African?" I don't think it's a good thing for people to attempt to derive any aspect of their identity from their genealogical ancestry, it has no bearing on who you are as a person, and has the potential for great harm. I've seen it used to give people a false sense of culture and belonging to a community that they previously had nothing to do with. If you felt a connection to a particular culture or identity, should we really be telling people that their genome is responsible for determining whether or not they can associate parts of those cultures with themselves? Racism results from a failure to separate identity from genome. If you listen to the dogma of a white supremacist, it's heavily predicated on associating the accomplishments of their "European" forefathers with themselves. If you listen to the dogma of a black supremacist, it's heavily predicated on assuming the identity of the Egyptians or Israelites and associating their accomplishments with themselves. I really think that if an individual is truly "disturbed" by not knowing the precise nature of their lineage, that is an individual in search of an identity on a plate, and therefore highly corruptible. The motivation to learn more about one's ancestry should ideally not surpass the level of an intellectual curiosity.
- devereaux 8y ago
  I wholeheartedly agree with your analysis on the risks. I still think it can have a positive role, if only to help the person emotional investment. > Is the usual "I'm part Irish part German part Cherokee" meaningfully more satisfying than "I'm west African?" For that, "I'm west African" is less satisfying than "I did trace my roots to Mali" which can lead the person to study the history of the country, Timbuktu, etc.
socrates1998 8y ago
These tests aren't a measure of ancestors, it's a measure of a person's DNA and where that DNA possibly comes from. Theoretically, they are similar, but since you inherit a random 50% of DNA from both parents, on an individual scale, you might inherit a misleading percent from your ancestors. Let's say 3 of your grandparents are all full-blooded Italian. Like, their families have been in the same village for centuries. And one of your grandparents is from England (and is similarly from an old family in a small village). So, theoretically your DNA results should 75% from that region of Italy and 25% from English region. However, lets say from your mom, the parent who is 50% English and 50% Italian, you only inherit a mix of 15% Italian and 35% English (15% + 35% = 50% mom's DNA). It's random, so this is possible, especially on a small sample of ONE person. And from your Dad, who is 100% Italian, you obviously inherit 50% of your DNA as from that Italian region. Now, you theoretically should be 75% Italian and 25% English, but since it's random, your true DNA reads as only 65% Italian and 35% English. This is why it's mostly bullshit to me. It's impossible to know what percents your DNA randomly passed on to your ancestors. Sure, in our example, the two places where correct, Italy and England, but the percents are way off. And that's honestly WHY we get these tests. We generally know where our families come from, we just have no idea what the percents are. So, the whole point of the service is stupid since they equate DNA with ancestry.
- devereaux 8y ago
  > We generally know where our families come from No, not everybody. Even percentage that are a bit off are better than nothing at all.
  - 0xffff2 8y ago
    Why though? By ancestry, I'm 25% Italian and a whole lot of other stuff. By percentage, I think I'm probably either mostly French or mostly English by ancestry, but that doesn't matter to me at all. What matters to me is that I grew up learning to cook from my Italian grandmother. As a result, in so far as I have any cultural identity other than "American", that identity is Italian, regardless of my genetic origins. It's hard for me to imagine why anyone would feel any significant connection to an ancestry that they needed a DNA test to know they had.

empath75 8y ago

I sent my grandparents raw dna file to a few different sites and got wildly different results, including 1% nigerian and 10% north african from myheritage.com (something that didn't show up at all on other sites.)