People all have prejudices. All people stereotype. Look at the old "Gilligan's Island" show. The professor, Russell Johnson, looks serious and smart and comes up with an invention virtually every show but still can't get the castaways off the island even though they're probably not far from civilization. Bob Denver as Gilligan looks like a hapless dimwit. The slovenly Captain looks like a jolly old fool. Mary Anne looks like a Kansas homecoming queen. Ginger looks like a low-end film star which is how Tina Louise's life ended up. The Howells look and talk rich. Perhaps if they had a black man in their three-hour tour expedition the castaways would have made it back to civilization but that would defeat the premise of the show.
This ties in with a comment I made from the other day about looks. It was a network news show that aired back in '85 about how looks shape our lives. I still remember being surprised at learning that teachers called on the better looking students more. And to your last question, it turned out that the "odd" looking students were filed into special education classes way more often than the more attractive students. The more attractive ones had to really show no proficiency to get moved into special ed. So, yes, people have strong stereotypes about what dumb and smart looks like. Most influential of all them all, as it turns out, may be teachers.
This assumes teachers call on students based on who they think smartest. I doubt that's the goal. Getting distributed participation is more likely. It also depends on whether they select from students raising their hands or make everyone do it.
We need a teach on this thread to tell us what the goal is.
At the extreme low end, sure a good looking student is less likely to get sent to the special classes. People are used to most retards being funny looking. It's probably rare to be good looking and poised yet retarded. In fact, without testing we'd probably never notice.
I'm wondering if any teacher out there would admit to calling on better looking students more often. I highly doubt it... even if true. It would make me look bad as a teacher.
Of course not. I meant to ask the teacher what the goal is of calling on students in the first place. It isn't immediately obvious to me why you would do that. That is, I can see why you would do it in a class discussion of a book but I don't understand why you would ask questions like the teacher in Ferris Bueler.
I have followed Dr Edward Dutton for years. One of his reccuring themes is that stereotypes about appearance are based on experiences and should not be dismissed out of hand. One of the stronger correlations is that symetry is associated with good genetics and intellegent people are more likely to be attractive than less intelligent ones all other things being equal. I myself shy away from these assertations out of political correctness and doubt that there are many researchers that do not do so to some extent. Meaning that the truth of these ideas will be slow in reaching the mainstream if they ever do.
Stand-up comics tend to split between acting dumber than they are in order to illuminate some absurdity or acting like the genius know-it-all who’s figured everything out and is sharing it with you. Who chooses which style doesn’t seem to have much to do with actual intelligence: I would put Norm Macdonald and Larry David in the first camp, Dave Chapelle, Jerry Seinfeld and John Mulaney in the second. Larry David in particular does a face to suggest he’s lowering his IQ by furrowing his brow and squinting. From SNL alumni I would say Mulaney and Bill Hader would struggle to be cast as a truly dumb character. Both are tall with angular facial features.
Teams like Allen and Rossi of the 1960s have an easier time of it. Marty Allen was goofy-looking with wild, frazzled hair and told the jokes. Steve Rossi was taller, angular and handsome. He was the straight man.
Much better nutrition and the general affluence of post-World War Two America propelled the rise of black dominance in basketball and, a little later, football. Also, the relative cheapness in building and maintaining basketball courts in cities launched the rapid dominance blacks obtained in basketball. Look on youtube of a mid-50s NBA championship basketball game. It's still a game of almost all whites. By 1965, the NBA was half-black. A majority of the best players were black. Bill Russell. Elgin Baylor. Wilt Chamberlain. Oscar Robertson. Sam Jones. Maurice Stokes.
Someone should go through all those "posture" photos from colleges in the 1940s to see how well people with odd bodies turned out compared to George HW Bush.
The UK, then about the richest country in the world, discovered in WW1 that an alarming number of her young men were malnourished and misshapen. Could it all have been because of cigarettes and gin?
Aristotle claimed that being attractive was one aspect of being a good person. He had a functionalist (or teleological) conception of human goodness, and, since being healthy is better for human beings than being unhealthy, being healthy is a part of what makes a human a good human. Likewise, for a variety of reasons including the ease with which attractive people can convince others and the pleasantness of being around attractive people, being attractive is part of what makes a human a good human. Being intelligent and wise are also important, but I don't think that Aristotle ever made the argument that these were logically or biologically connected, but were instead practically mutually supporting. On the other hand, Plato made the argument that attractive people were more susceptible to flattery and thus more likely to go off the rails (and suggest that Athens invade Sicily or some similarly silly idea).
My take on shaved heads on white guys is that it's much cheaper and more effective to overcome male pattern baldness with a razor than with Rogaine. It doesn't hurt that a shaved head has a certain 'bad-ass' stereotype, either.
Definately cheaper and easier to shave the head than get regular haircuts. A freshly shaved skull is much better looking than a bad or badly overdue haircut.
I agree with Christopher B, JMcG, & Kelly Harbeson. The "badass" impression still exists but has declined into more like the head-shaver is not habitual conformist, although maybe some of us do look to some folks like we are not to be provoked.
However, in my case yet another reason to shave exists. My remaining hair is difficult to manage and must be cut short, while hair cutters are trying to charge for styling and expect a tip. Hence, I am offended by the pricing for a very simple job.
I'm going to read that "missing heritability" article (it's very long) but here is my initial thoughts: In medicine you disbelieve what the micro results suggest, if the macro results contradict them. Humans are complex systems with unpredictable emergent properties. If you find a gene that should stop cancer in it's tracks if you turn it off, and you turn of the gene, and the cancer deaths barely change, you don't say the cancer is wrong. Your hypothesis based on the gene was wrong.
It's entirely possible that the twin studies got it wrong. The idea that it's based on looks is stupid beyond belief. I can't believe we would even discuss that seriously. I could go into personal experience in detail if anyone needs it.
Back to the GWAS studies. My educated guess? Humans are complex systems and we shouldn't expect our puny minds and deterministic thinking to be able to make predictions. Anyone who was surprised by chatGPT, i.e. the emergent properties of next word prediction at scale, needs to admit that we don't know what the fuck we are doing with complex systems. I was shocked by it. All of you were shocked by it. The only people not shocked by it are the ones who don't know how it works and just treat all tech like magic.
GWAS studies are based on SNPs. They don't tell you direct information about functional gene variants. I haven't kept up on the literature so I concede when you do that many. it's possible they are drilling down on specific variants, but I still think that many different SNP patterns could have the exact same variant or functionally same variant of any given gene.
So you have a million plus SNPs some of which indicate different genes some of which don't (and I mean person to person different or not). Some of these will indicate variants in proteins, some in regulators, some in who knows what.
Then you take a sample size of fewer people than we have actual genes, and you expect all the complex interactions to show signal in a deterministic manner and predict a single trait?
Anyway, I think the problem is in there somewhere. That doesn't mean that IQ is not mostly inherited.
Also, we might be making a bigger deal out of marginal differences than we should. We can look for height genes but as we do, we should consider that we make a very big deal about the difference between 5'8" and 6'2" while in the context of all mammals heights that's a rounding error.
Intelligence is almost certainly controlled by more than one gene. Probably lots. I doubt they’ve all been identified.
I also doubt much of a connection with looks: the exception being genetic conditions that affect both looks and intelligence. Down syndrome comes to mind.
But it’s not caused by a single gene but by an extra copy of chromosome 21.
More importantly, looks and intelligence are similar in that they are likely mostly the product of developmental and embryological instructions, as opposed to ongoing protein manufacture and regulation.
The model that most people appear to have, though they don't say it, is that higher intelligence is due to the sum of a million ever so slightly better genes for intelligence. But what would those genes be? Are the sodium channels in your neurons a little faster? Do your vesicles release neurotransmitter into the synapse slightly faster because you have "better proteins?"
or is it more matter of brain architecture, the shape, the connections? I figure it's mostly the latter and that means that the "better" intelligence genes are summing into sequential and parallel instructions that make for a better architecture. I don't think we are close to knowing how that works.
But it is similar to a better face. You don't have a better face because you have better genes for, I dunno, keratin. You have a better face because from embryology up through development, the instructions made your face grow in a pleasing manner.
Once that better looking architecture is set up by late adolescence, there's no ongoing protein regulation that keeps you good looking. All there is, is the ravages of time tearing it down :(
oh- and your nose and ears keep growing for some reason.
It seems quite plausible that people with certain appearances are considered good looking partly because we were selected to be pleased by certain features of appearance. Part of that might be indirect selection for lack of genetic defects.That selection would be by community as well as by "survival of the fittest" (individual selection) for favoring appearance traits. Mostly, it would be selection by community.
I'm sorry, but I did not know what "GWAS" and "SNP" stand for in this context, including the article and all the comments. I was hoping to come across a comment that would directly or indirectly define the abbreviations. This is a link to a verbose explanation that GWAS is "genome-wide association study", which means the entire genome is somehow classified and the categories of entire genome patterns are used to try to see what causes physical traits or diseases. Being government product, the definition offered is verbose.
SNP=single nucleotide polymorphism. It's a single base pair change that is silent (i.e. it doesn't change the protein coded for). The importance in the context of a genome wide association study is that an SNP does affect whether or not a restriction enzyme will cut the genome at that point. GWAS are not performed by sequencing entire genomes and comparing them. That is still too expensive (last time I checked which was a decade ago:) ).
Instead you slice up the genome with a bunch of restriction enzymes and count on patterns in the fragments you see (which vary based on the different SNPs individuals inherited) to judge whether people inherited the same genes.
So it's not comparing genes directly across populations. it's looking for patterns in SNPs (indirectly) and hoping that those patterns of SNP are associated with (on average) common genes.
Thanks! But here is more trouble to handle. :) The government web page says that the GWAS specimens are put into multiple regression analyses. Multiple regression requires -- among other things -- numerical values spaced at equal distances and having a zero point. Some violation can be justified, such as in measuring samples of gaseous mass or human body weight, where the zero point means there is no specimen. In social sciences, this measurement-level requirement is simply ignored. So how are such values assigned to genome-wide specimens or to snips of genomes? Or is the problem just ignored, as is done is sociology?
I don't know how the stats are done on a GWAS study. I suppose I assumed it was machine learning of some kind. I suppose if they really are doing a traditional linear regression analysis they might be doing it one SNP and one trait at a time. Like they could take a set of people who had heart attacks before age 50 and those who did not and compare the prevalence of. . . no I don't get it. I cannot immediately intuit it. given the number of human traits they might look for and the huge number of SNPs and the (as it turns out) small amount of signal from each SNP on each trait, I think you need to do some kind of machine learning approach.
I'd look it up, but my impression is that so far the results have been almost universally disappointing. That's cool information too but it makes me less motivated to criticize their statistical methods.
I wonder if people are confusing (a) the subjective experience of perceiving discrimination (not great!) with (b) how that perceived discrimination objectively affects someone’s life. For example, take the Henry Louis Gates episode during Obama’s presidency. He seems like he genuinely felt bad about being questioned by the police outside his home. But, and not to undermine how bad he felt in the moment, the episode was good for him, from the perspective of an observer, because it raised his profile and he was able to do his tv show. To your post, it seems like researchers and their audience want to show that discrimination is a big deal, because discrimination, like rejection of any kind, feels bad. But if people are resilient, and they can find alternative opportunities, then that discrimination isn’t going to show up in objective measures, even if it’s 100% to feel like shit when you think you’ve been treated unfairly.
It is a well known problem that GWAS represents heredity "poorly"*. The simple answer is that GWAS is using simple multi-linear regression to keep the number of degrees of freedom down. Say you have 0, 1, or 2 copies of variant Alpha for Gene 1. Simplest way of assessing effect is linear. Obviously, this will produce problems when variant Alpha is recessive - then you have a phenotype cut-off between 1 and 2 copies, or if Alpha is dominant - then you have a phenotype cut-off between 0 and 1 copies.
But that is not the end of it. Let's say 1 copy of variant Alpha and 1 copy of variant Omega for the same gene 1 produce same effect as 2 copies of recessive variant Alpha? Or variants Apha and Omega are dominant, and it's now 1 copy of either one that is sufficient to achieve phenotype? So two is a without effect surplus?
And then you can have combinations between different genes, say gene 1 and gene 357.
The problem is that one could start accounting for these combinations. However, it would lead to an explosion in the number of variational degrees of freedom that would tend to kill any multi-linear regression statistic. Whereas studies with separately adopted identical twins do not suffer from any such concern.
By the way, with genes I don't just mean proteins and variations in the exon DNA-encoded aminoacid sequence. All the promoter and splicing information in the non-coding DNA is just as important. And much of it is mutating quite fast.
SO MY MAIN POINT IS THAT FOR GWAS MULTILINEAR REGRESSION IS A NECESSITY TO KEEP DEGREES OF FREEDOM DOWN, BUT IT IS SUBOPTIMAL FOR THE EXISTING "GENETIC" ALGEBRA WHICH OPERATES WITH A COMBINATION OF LINEAR ELEMENTS AND COMPLEX LOGIC APPARATUS.
*If heredity is responsible for 75% of the variance, but GWAS can only account for 15% of the variance, ie only 20% of the whole hereditable variance, despite samples of 5Mio size, this is what I would call "poorly".
Laurence Fox is both highly intelligent and attractive (though that may be my Anglophilia for crime serials, as I also found late-stage John Thaw attractive). I love any post that gives Tracy Morgan his due. Remember when Tina Fey had the gall to call out Morgan in a particularly humiliating way for his purported homophobia? Because he made a joke? Remember when Stephen Colbert got his start viciously stereotyping gays for years on Strangers With Candy (a fine show)?
'Do people really have strong stereotypes about how smart and dumb people look?'
Don't people have the stereotype that really attractive women are not smart?
And that the smartest guys are the nerdiest looking ones?
If how people perceive your looks is what affects your IQ, then Jerry Lewis should in reality have been dumber than Dean Martin.
One thing I noticed in a third year physics class of about 20 people (I was the only woman): all of us (including the prof) were near-sighted.
That seems like more than a coincidence.
Just my own experience but I believe that a lot of reading, especially under less than optimal lighting tends to foster nearsightedness.
There’s a strong hereditary component but reading may also cause it.
https://medlineplus.gov/genetics/condition/nearsightedness/#causes
I would surmise that there is a large hereditary component to a tendency to alot of reading.
It’s mostly hereditary. My mother was near-sighted. My daughter is too.
Reading can contribute to it, according to the article.
People all have prejudices. All people stereotype. Look at the old "Gilligan's Island" show. The professor, Russell Johnson, looks serious and smart and comes up with an invention virtually every show but still can't get the castaways off the island even though they're probably not far from civilization. Bob Denver as Gilligan looks like a hapless dimwit. The slovenly Captain looks like a jolly old fool. Mary Anne looks like a Kansas homecoming queen. Ginger looks like a low-end film star which is how Tina Louise's life ended up. The Howells look and talk rich. Perhaps if they had a black man in their three-hour tour expedition the castaways would have made it back to civilization but that would defeat the premise of the show.
This ties in with a comment I made from the other day about looks. It was a network news show that aired back in '85 about how looks shape our lives. I still remember being surprised at learning that teachers called on the better looking students more. And to your last question, it turned out that the "odd" looking students were filed into special education classes way more often than the more attractive students. The more attractive ones had to really show no proficiency to get moved into special ed. So, yes, people have strong stereotypes about what dumb and smart looks like. Most influential of all them all, as it turns out, may be teachers.
This assumes teachers call on students based on who they think smartest. I doubt that's the goal. Getting distributed participation is more likely. It also depends on whether they select from students raising their hands or make everyone do it.
We need a teach on this thread to tell us what the goal is.
At the extreme low end, sure a good looking student is less likely to get sent to the special classes. People are used to most retards being funny looking. It's probably rare to be good looking and poised yet retarded. In fact, without testing we'd probably never notice.
I'm wondering if any teacher out there would admit to calling on better looking students more often. I highly doubt it... even if true. It would make me look bad as a teacher.
Of course not. I meant to ask the teacher what the goal is of calling on students in the first place. It isn't immediately obvious to me why you would do that. That is, I can see why you would do it in a class discussion of a book but I don't understand why you would ask questions like the teacher in Ferris Bueler.
I have followed Dr Edward Dutton for years. One of his reccuring themes is that stereotypes about appearance are based on experiences and should not be dismissed out of hand. One of the stronger correlations is that symetry is associated with good genetics and intellegent people are more likely to be attractive than less intelligent ones all other things being equal. I myself shy away from these assertations out of political correctness and doubt that there are many researchers that do not do so to some extent. Meaning that the truth of these ideas will be slow in reaching the mainstream if they ever do.
Stand-up comics tend to split between acting dumber than they are in order to illuminate some absurdity or acting like the genius know-it-all who’s figured everything out and is sharing it with you. Who chooses which style doesn’t seem to have much to do with actual intelligence: I would put Norm Macdonald and Larry David in the first camp, Dave Chapelle, Jerry Seinfeld and John Mulaney in the second. Larry David in particular does a face to suggest he’s lowering his IQ by furrowing his brow and squinting. From SNL alumni I would say Mulaney and Bill Hader would struggle to be cast as a truly dumb character. Both are tall with angular facial features.
Teams like Allen and Rossi of the 1960s have an easier time of it. Marty Allen was goofy-looking with wild, frazzled hair and told the jokes. Steve Rossi was taller, angular and handsome. He was the straight man.
Of those I would guess Norm to be the one with super high intelligence with John Mulaney well above average.
Norm almost won a million on “Who wants to be a millionaire?” until Regis Philbin led him astray on the very last question.
Much better nutrition and the general affluence of post-World War Two America propelled the rise of black dominance in basketball and, a little later, football. Also, the relative cheapness in building and maintaining basketball courts in cities launched the rapid dominance blacks obtained in basketball. Look on youtube of a mid-50s NBA championship basketball game. It's still a game of almost all whites. By 1965, the NBA was half-black. A majority of the best players were black. Bill Russell. Elgin Baylor. Wilt Chamberlain. Oscar Robertson. Sam Jones. Maurice Stokes.
Someone should go through all those "posture" photos from colleges in the 1940s to see how well people with odd bodies turned out compared to George HW Bush.
The UK, then about the richest country in the world, discovered in WW1 that an alarming number of her young men were malnourished and misshapen. Could it all have been because of cigarettes and gin?
“Missing Maleability” is a brilliant turn of phrase
Aristotle claimed that being attractive was one aspect of being a good person. He had a functionalist (or teleological) conception of human goodness, and, since being healthy is better for human beings than being unhealthy, being healthy is a part of what makes a human a good human. Likewise, for a variety of reasons including the ease with which attractive people can convince others and the pleasantness of being around attractive people, being attractive is part of what makes a human a good human. Being intelligent and wise are also important, but I don't think that Aristotle ever made the argument that these were logically or biologically connected, but were instead practically mutually supporting. On the other hand, Plato made the argument that attractive people were more susceptible to flattery and thus more likely to go off the rails (and suggest that Athens invade Sicily or some similarly silly idea).
Kelsey Grammer looks intelligent. Whoever was casting director for Cheers knew what she was doing.
A certain kind of male pattern baldness seems to be a visual shorthand for smarts. Why?
Does it imply the wisdom of old age? It seems like half the men in the world shave their heads and grow beards now.
My take on shaved heads on white guys is that it's much cheaper and more effective to overcome male pattern baldness with a razor than with Rogaine. It doesn't hurt that a shaved head has a certain 'bad-ass' stereotype, either.
It did, but it’s so common now that I don’t think that’s still true. Exhibit A: Jeff Bezos.
Definately cheaper and easier to shave the head than get regular haircuts. A freshly shaved skull is much better looking than a bad or badly overdue haircut.
Blumenbach thought that people with higher brows were literally more intelligent: hence “highbrow” or “egghead”.
I agree with Christopher B, JMcG, & Kelly Harbeson. The "badass" impression still exists but has declined into more like the head-shaver is not habitual conformist, although maybe some of us do look to some folks like we are not to be provoked.
However, in my case yet another reason to shave exists. My remaining hair is difficult to manage and must be cut short, while hair cutters are trying to charge for styling and expect a tip. Hence, I am offended by the pricing for a very simple job.
I'm going to read that "missing heritability" article (it's very long) but here is my initial thoughts: In medicine you disbelieve what the micro results suggest, if the macro results contradict them. Humans are complex systems with unpredictable emergent properties. If you find a gene that should stop cancer in it's tracks if you turn it off, and you turn of the gene, and the cancer deaths barely change, you don't say the cancer is wrong. Your hypothesis based on the gene was wrong.
It's entirely possible that the twin studies got it wrong. The idea that it's based on looks is stupid beyond belief. I can't believe we would even discuss that seriously. I could go into personal experience in detail if anyone needs it.
Back to the GWAS studies. My educated guess? Humans are complex systems and we shouldn't expect our puny minds and deterministic thinking to be able to make predictions. Anyone who was surprised by chatGPT, i.e. the emergent properties of next word prediction at scale, needs to admit that we don't know what the fuck we are doing with complex systems. I was shocked by it. All of you were shocked by it. The only people not shocked by it are the ones who don't know how it works and just treat all tech like magic.
GWAS studies are based on SNPs. They don't tell you direct information about functional gene variants. I haven't kept up on the literature so I concede when you do that many. it's possible they are drilling down on specific variants, but I still think that many different SNP patterns could have the exact same variant or functionally same variant of any given gene.
So you have a million plus SNPs some of which indicate different genes some of which don't (and I mean person to person different or not). Some of these will indicate variants in proteins, some in regulators, some in who knows what.
Then you take a sample size of fewer people than we have actual genes, and you expect all the complex interactions to show signal in a deterministic manner and predict a single trait?
Anyway, I think the problem is in there somewhere. That doesn't mean that IQ is not mostly inherited.
Also, we might be making a bigger deal out of marginal differences than we should. We can look for height genes but as we do, we should consider that we make a very big deal about the difference between 5'8" and 6'2" while in the context of all mammals heights that's a rounding error.
Intelligence is almost certainly controlled by more than one gene. Probably lots. I doubt they’ve all been identified.
I also doubt much of a connection with looks: the exception being genetic conditions that affect both looks and intelligence. Down syndrome comes to mind.
But it’s not caused by a single gene but by an extra copy of chromosome 21.
More importantly, looks and intelligence are similar in that they are likely mostly the product of developmental and embryological instructions, as opposed to ongoing protein manufacture and regulation.
The model that most people appear to have, though they don't say it, is that higher intelligence is due to the sum of a million ever so slightly better genes for intelligence. But what would those genes be? Are the sodium channels in your neurons a little faster? Do your vesicles release neurotransmitter into the synapse slightly faster because you have "better proteins?"
or is it more matter of brain architecture, the shape, the connections? I figure it's mostly the latter and that means that the "better" intelligence genes are summing into sequential and parallel instructions that make for a better architecture. I don't think we are close to knowing how that works.
But it is similar to a better face. You don't have a better face because you have better genes for, I dunno, keratin. You have a better face because from embryology up through development, the instructions made your face grow in a pleasing manner.
Once that better looking architecture is set up by late adolescence, there's no ongoing protein regulation that keeps you good looking. All there is, is the ravages of time tearing it down :(
oh- and your nose and ears keep growing for some reason.
It seems quite plausible that people with certain appearances are considered good looking partly because we were selected to be pleased by certain features of appearance. Part of that might be indirect selection for lack of genetic defects.That selection would be by community as well as by "survival of the fittest" (individual selection) for favoring appearance traits. Mostly, it would be selection by community.
I'm sorry, but I did not know what "GWAS" and "SNP" stand for in this context, including the article and all the comments. I was hoping to come across a comment that would directly or indirectly define the abbreviations. This is a link to a verbose explanation that GWAS is "genome-wide association study", which means the entire genome is somehow classified and the categories of entire genome patterns are used to try to see what causes physical traits or diseases. Being government product, the definition offered is verbose.
https://www.genome.gov/genetics-glossary/Genome-Wide-Association-Studies-GWAS
SNP=single nucleotide polymorphism. It's a single base pair change that is silent (i.e. it doesn't change the protein coded for). The importance in the context of a genome wide association study is that an SNP does affect whether or not a restriction enzyme will cut the genome at that point. GWAS are not performed by sequencing entire genomes and comparing them. That is still too expensive (last time I checked which was a decade ago:) ).
Instead you slice up the genome with a bunch of restriction enzymes and count on patterns in the fragments you see (which vary based on the different SNPs individuals inherited) to judge whether people inherited the same genes.
So it's not comparing genes directly across populations. it's looking for patterns in SNPs (indirectly) and hoping that those patterns of SNP are associated with (on average) common genes.
Thanks! But here is more trouble to handle. :) The government web page says that the GWAS specimens are put into multiple regression analyses. Multiple regression requires -- among other things -- numerical values spaced at equal distances and having a zero point. Some violation can be justified, such as in measuring samples of gaseous mass or human body weight, where the zero point means there is no specimen. In social sciences, this measurement-level requirement is simply ignored. So how are such values assigned to genome-wide specimens or to snips of genomes? Or is the problem just ignored, as is done is sociology?
I don't know how the stats are done on a GWAS study. I suppose I assumed it was machine learning of some kind. I suppose if they really are doing a traditional linear regression analysis they might be doing it one SNP and one trait at a time. Like they could take a set of people who had heart attacks before age 50 and those who did not and compare the prevalence of. . . no I don't get it. I cannot immediately intuit it. given the number of human traits they might look for and the huge number of SNPs and the (as it turns out) small amount of signal from each SNP on each trait, I think you need to do some kind of machine learning approach.
I'd look it up, but my impression is that so far the results have been almost universally disappointing. That's cool information too but it makes me less motivated to criticize their statistical methods.
I've always had the impression that smart people are more homely (and more overweight, and more myopic) than dumb people.
Present company excepted.
I’ve definitely noticed a correlation between near-sightedness and ability to study physics.
I wonder if people are confusing (a) the subjective experience of perceiving discrimination (not great!) with (b) how that perceived discrimination objectively affects someone’s life. For example, take the Henry Louis Gates episode during Obama’s presidency. He seems like he genuinely felt bad about being questioned by the police outside his home. But, and not to undermine how bad he felt in the moment, the episode was good for him, from the perspective of an observer, because it raised his profile and he was able to do his tv show. To your post, it seems like researchers and their audience want to show that discrimination is a big deal, because discrimination, like rejection of any kind, feels bad. But if people are resilient, and they can find alternative opportunities, then that discrimination isn’t going to show up in objective measures, even if it’s 100% to feel like shit when you think you’ve been treated unfairly.
It is a well known problem that GWAS represents heredity "poorly"*. The simple answer is that GWAS is using simple multi-linear regression to keep the number of degrees of freedom down. Say you have 0, 1, or 2 copies of variant Alpha for Gene 1. Simplest way of assessing effect is linear. Obviously, this will produce problems when variant Alpha is recessive - then you have a phenotype cut-off between 1 and 2 copies, or if Alpha is dominant - then you have a phenotype cut-off between 0 and 1 copies.
But that is not the end of it. Let's say 1 copy of variant Alpha and 1 copy of variant Omega for the same gene 1 produce same effect as 2 copies of recessive variant Alpha? Or variants Apha and Omega are dominant, and it's now 1 copy of either one that is sufficient to achieve phenotype? So two is a without effect surplus?
And then you can have combinations between different genes, say gene 1 and gene 357.
The problem is that one could start accounting for these combinations. However, it would lead to an explosion in the number of variational degrees of freedom that would tend to kill any multi-linear regression statistic. Whereas studies with separately adopted identical twins do not suffer from any such concern.
By the way, with genes I don't just mean proteins and variations in the exon DNA-encoded aminoacid sequence. All the promoter and splicing information in the non-coding DNA is just as important. And much of it is mutating quite fast.
SO MY MAIN POINT IS THAT FOR GWAS MULTILINEAR REGRESSION IS A NECESSITY TO KEEP DEGREES OF FREEDOM DOWN, BUT IT IS SUBOPTIMAL FOR THE EXISTING "GENETIC" ALGEBRA WHICH OPERATES WITH A COMBINATION OF LINEAR ELEMENTS AND COMPLEX LOGIC APPARATUS.
*If heredity is responsible for 75% of the variance, but GWAS can only account for 15% of the variance, ie only 20% of the whole hereditable variance, despite samples of 5Mio size, this is what I would call "poorly".
Laurence Fox is both highly intelligent and attractive (though that may be my Anglophilia for crime serials, as I also found late-stage John Thaw attractive). I love any post that gives Tracy Morgan his due. Remember when Tina Fey had the gall to call out Morgan in a particularly humiliating way for his purported homophobia? Because he made a joke? Remember when Stephen Colbert got his start viciously stereotyping gays for years on Strangers With Candy (a fine show)?