Feedback

Hi all,

Please leave some suggestions below on how this site could be improved. Any ideas/criticisms are welcome.

1,888 Responses

  1. passingtime85 says:

    I’m going to attempt to break this comment up, so the comment section of the Dora Madison Burge page doesn’t get clogged. I tried posting it once, but I think it was automatically counted as spam and was deleted.

    @Andrew133

    https://www.reddit.com/r/AncestryDNA/comments/nphl94/common_traits_for_this_ethnicity/

    First things first, those are old results, secondly, those could be her initial results, which tend to be inaccurate and updates tweak the results over time.

    Thirdly AncestryDna isn’t all that accurate.

    Fourthly no commercial company accounts for ancient admixture events. All their data is based on reference populations of people that can prove their origins, several generations back, specifically linking them to a geographic location.

    • passingtime85 says:

      That’s like if researchers 10 centuries from now, were to take samples from UK citizens with roots in India, and then actually call them indigenous to Britain. Then justify that assertion by stating, the bloodlines had been there for so long, they’ve effectively become British.

      Which is true under a certain point of view. They would be distinctive from proper Indian populations. They probably would have new haplogroups not found in India, and they’d also undoubtedly have admixture from other ethnic groups. But they wouldn’t truly be British as we’d think of them today. They’d be a unique new group, but lazy researchers would just call them British, because of the duration of their presence on the island. Instead of calling them some new, more accurate moniker, like Indo-Brits or something of that ilk.

      The same thing happens when you presently look at people that genetically skew local, to any given area, but look phenotypically look anomalous, atypical, uncommon, etc.

      • passingtime85 says:

        You can look at autosomal test results and say “hey the sample compared to the reference population matches, so that means this person is indeed from area X”. Then you look at the person and say, “then why does the person look so out of place compared to their geographic cousins?”

        Two explanations or reasons are among the most plausible.

        Reason 1.

        You simply haven’t been exposed to the wide variety of variation in that area. You could see tens to hundreds of thousands of people in any location, but that isn’t a complete representation, of that area’s population. Even if you’re looking at samples from border to border, or from each corner of a region, you still may not see millions of individuals, all of which may not fit your mental image of the phenotypic homogeny.

        Reason 2.

        The reference sets don’t account for ancient admixture events.

        • passingtime85 says:

          We’ve talked about this before.

          As an example, Ancient North East Asians effected Scandinavians, which trickles down to all the descendants of the Scandinavians themselves, across Europe. Ancient North East Asians and Ancient Siberians also have ancient roots from Europe, which effects all their descendants as well, including those people that settled in the Americas. These are just two examples, there are many, many more.

          Human history, migration, genetic exchange, genetic drift, and admixture events are all sloppy. People ebb and flow into areas and leave traces of themselves in populations that remain. This happens everywhere where humans exist. Through haplogroups, through gene markers, through physical traits, etc.

          It is however, very possible to recognize clines in groups, frequencies, and typical features.

          • passingtime85 says:

            When something or someone looks out of place, there’s usually a reason for it, if you look deeply enough into the matter.

            Science isn’t magic. It’s observation, recording, hypothesis, experimentation, findings, and theory. Those theories can always be altered, when new observations and new variables are tested against those theories.

            The more information you have, the more accurate the research becomes, which hones our understanding of observable phenomena.

            We all have to recognize genetic research is an ever evolving field of study. In part because there are incomplete data sets, lost to the ravages of time. Another reason is that the modern world has more admixture events than ever, because of ease of transportation, so groups are are in constant flux.

          • passingtime85 says:

            Not to mention the apprehension of researchers to publish and/or promote ideas, they view are detrimental to society as a whole. I.E. they don’t want to promote what they consider bias, through the findings of scientific study. The general population that aren’t in the lab, has to take a larger view of these findings, and decide for themselves what is closest to being accurate.

            Researchers are not divine. They are people with opinions and agendas of their own. They will represent their findings according to how they believe the data should be interpreted, and you are allowed to disagree with their findings. Don’t treat scientists as the highest authority of the land. That’s what billions of people have done for ages, when they look to theologians and clergy for their answers. You’re trading one religion for another.

            Look over the research yourself and learn how interpret the findings. Don’t always allow the scientists to spoonfeed you the answers, while assuming what they tell you, is everything there is to say upon the matter.

          • passingtime85 says:

            Sometimes they leave things out, because they think certain information is inconsequential to the whole, or because they think it’s detrimental to the understanding of the topic. That way of controlling or presenting the description phenomena, is myopic. To understand anything fully, or even just better/more, you have to look at the broader scope.

            Also remember they’re fallible, even when a published study, is peer reviewed, it still can be fallible. It’ll be accepted for a while, until someone debunks it.

            There have been numerous studies on the accuracy of peer review study. Even when researchers have their peers review their work, their research is still not particularly reliable or accurate, as a predictive model. The act of peer review itself, regardless of content or field, has even been studied.

            This has been statistical studied over and over again, and honestly people make mistakes all the time, and it takes a ton of work, to refute each screw up. Many papers get retracted from national science journals, very frequently, even after they were “peer reviewed” and published.

          • andrew says:

            Nevermind, but what do you personally think about the young girl from Reddit and Mrs. Burge.

        • passingtime85 says:

          They’ve done psychological tests on this. One study went like this. A researcher sent out a paper to 35 individual researchers, that had they themselves researched the same topic in the past. The paper was judging the efficacy of transcutaneous electrical nerve stimulation. The paper was basically made up, and all the copies were the same except the conclusion/findings. Half showed the therapy in a positive light, half in a negative. When people disagreed with the findings more, they harshly criticized the methodology, even though, there was no difference. The people were just being biased off their own preconceived notions.

          • passingtime85 says:

            There’s been other studies like sending out a paper done on the efficacy of social work, and again the same thing happened. If the conclusion was favorable, more people were willing to publish the work, if it was negative more people were of the mind to decline publishing the work, even though the paper was always the same, just the conclusion was different.

            There was another bogus study sent out for review discussing joint surgery, half were positive, the other half was not negative, but they showed no benefit to the surgery. It had 7 intentional errors in it. People that wanted to publish the article found less than half of 1 error, even though they we’re obvious and intentional. The people that denied publication only found 1 error. 97% of the people given the paper with positive results gave a vote for publication. 80% gave a judgment for denial of publication that were given the paper that showed no benefit. People in favor of surgery rated the methodology higher, than the people that received the paper that didn’t see the benefit of surgery.

            Study after study shows these researchers can be wrong and not alot happens push back wise. People support what they support and deny what opposes their views, just out of instinct, without really delving into the reasoning. They don’t even really take these things, their assertions, and what they promote, incredibly seriously.

          • passingtime85 says:

            There was a study that tracked the acceptance or rejection of tens of thousands of articles submitted for review by the American Heart Association journal. If an author worked for a major institution like a university, company, hospital, their peers were more likely to give their research paper a stamp of approval, at a rate of 57%.

            In an semi related, but different, study, a double double blind study was conducted. Where the reviewer was not told the name of the author. Most reviewers can guess about 38% accuracy who is trying to get the paper published, by the writing style and content of the article. If they have a favorable review of the author, again they’re still more likely to promote the paper. So just name recognition can get you published and stamped as being vetted by your peers.

            People’s preferences and bias towards any topic, and their promotion of their pov, is outlandish, and even more outlandish is their suppression of opposing view points.

          • passingtime85 says:

            This doesn’t just pertain to one type of research, it bleeds over to every other field of study. When it comes to biology, microbiology, medicine, ecology, geology, meteorology, virology, epidemiology, bacteriology, pharmacology, genetics etc etc. basically the study of every facet of the natural world, there are alternating and opposing theories and/or interpretations of data.

            The schisms extend to every field of study known to mankind, there are no absolutes. Various concepts in math are in contention, so too in the fields of physics, technology, philosophy, epistemology, on and on.

            There’s almost no concrete, solidified, unified consensus and/or understanding of anything, because information is always being added, to every area of study. Only mathematical proofs are accepted as irrefutable, but there’s people out there that actually try to refute them, through rigorous testing and upper levels of mathematical disciplines. Lots of theorems hold true for the most part, until something is observed, that changes the understanding of the subject in question.

          • passingtime85 says:

            But like I said there are clines, generalities that can be representative of the common, frequently occurring phenomenon. In arithmetic 1+1=2, but in other mathematical disciplines it can actually be something else.

            99.9999% of the time however, the answer is just 2. That might be myopic on my part, but for general everyday life, it’s well understood and well accepted.

            So when you look to those scenarios where the answer, or reason, for a discrepancy in the majority vs the minority, it is in fact, sometimes completely alright to generalize, or use common findings, reasoning, or logic.

            In the question of ‘why is this out of place?’. You can use Occam’s razor, because it came from somewhere else.

            Or more specifically to our area of interest: In the case of genealogical study, in relation to genetic drift, and how it relates to phenotypic traits and relationship to ancestral region affinity/origin, and why some people don’t resemble their supposed regional kin?

          • passingtime85 says:

            The data and/or results of autosomal testing, does not explain or account for, the complete chronicle of ancestral lineage, that we have, thus far, recorded, observed, or studied.

            Generalized traits morph as you trace your finger along a map, in a gradient fashion. There will be pockets of anomalous features, and almost all of those can either be attributed to isolation, endogamy, or an admixture event, or several events.

            If someone looks odd or out of place, maybe it’s because you need to expand your parameters of regional heritage/racial/ethnic classification, or just as likely, or possibly more, because they have ancestry that hails from a foreign location.

  2. momof3 says:

    All of the celebrities that are tagged with “Kannada” should be changed to “Kannadiga” since Kannada is the language, and Kannadiga is the ethnicity. The change should also be made to the ethnicity section on the actual page.

  3. passingtime85 says:

    Is muricans still on here? I think an exchange with automatik would provide a jaunty read.

  4. passingtime85 says:

    Dang, bearboy (sorry I prefer this handle the most) really gets censored often these days. Is it because he gets political? I mean, tackled the unrest in the ME and that exchange is still up…Is it because of his criticism of a sitting president? Or a Democratic one, or one closing in on the end of lucidity?

    Hadnodiner is a bigger troll than bearboy, but his unnecessary comments are tolerated. Just a curiosity.

    • jackson9 says:

      Obviously because I said something negative of a sitting Democrat president. If it’s trump the admins would defend comparing him to the nazi regime. Which one did when user Neil Tennant did do.

      • jackson9 says:

        I’ll make a clear question for an admin to answer. Why when Neil Tennant called Trump a Nazi not only was it kept up but the language was justified by @Follers. But when I question Biden’s faith it gets deleted? @Follers you have the floor as to why I am censored and Neil Tennant’s worse accusations weren’t? Please explain to us why this isn’t hypocrisy?

  5. MichaelWHat says:

    Ethnic was it you who emailed me on the 16th about an internship with your website or was that spam?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.