Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Genetic data from over 20k U.S. children misused for 'race science'

nytimes.com

Post reply on HN
Print view View original

31–40 of 88 posts

Re: Genetic data from over 20k U.S. children misused for 'race science'

#31

Post by terminalshort »

yed wrote: post #21
terminalshort wrote: post #16

Earlier quoted context omitted.

OK, then let's do it right. But I think you know that isn't really the issue here. Nobody is putting out studies on correctly defined races by genetic groupings and intelligence either because the topic is still considered heresy. Your point that the commonly used definition of race is inaccurate is simply deflecting from this fact.

The point is there is no such thing as a “correct” grouping. The choice of what constitutes a group is completely subjective, it all depends on how far you choose to zoom in or out.

There absolutely are correct and incorrect ways of partitioning data into groups. Human genetics is no exception.

https://en.wikipedia.org/wiki/Cluster_analysis

https://en.wikipedia.org/wiki/Haplogroup

Re: Genetic data from over 20k U.S. children misused for 'race science'

#32

Post by dismalaf »

> Jordan Lasker, who often writes about race and intelligence under the name Crémieux

It's funny because this guy is center-left, he just happens to actually be intellectually honest.

Anyhow, either we do science or we just admit that we don't like the social implications of the evidence. Trying to hide data and gaslight the public isn't science.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#33

Post by terminalshort »

dragonwriter wrote: post #30
terminalshort wrote: post #12

Earlier quoted context omitted.

You are a perfect example of what I am talking about. That is an obvious lie. Of course it means something scientifically because it refers to a genetic grouping. e.g. black people are much more likely to have the genetic disease sickle cell anemia. Nobody seems to have much of a problem with that statement but if it were a gene related to intelligence you people would lose your minds. If it isn't a scientifically va…

Races are not genetic groupings, they are social constructs whose boundaries evolve over time, which is particularly clear when they are formalized in a way which resista change and that formalization drifts increasingly far from the current common usage, such as the way the White racial category in common usage in America currently roughly corresponds to the the subset of the White racial category that excludes the…

You are 100% correct, but also 100% missing the point. When I say white people are more likely to carry the gene for cystic fibrosis, or black people are more likely to carry genetic risk factors for kidney disease, nobody will reply with a long winded explanation claiming that statement is invalid because "white" and "black" are not scientifically valid because... Comments like yours appear only when the topic of intelligence comes up, so I conclude that the real problem you have with this is the subject of intelligence, and not the categories.

And when the science on race and intelligence came out, the response of the scientific community was not "your categories are bad, and here is my study on intelligence that actually uses scientifically valid genetic groupings." It was "any further science on this subject will not be funded and if you express disagreement it will risk your career."

Re: Genetic data from over 20k U.S. children misused for 'race science'

#34

Post by yed »

terminalshort wrote: post #31
yed wrote: post #21

Earlier quoted context omitted.

The point is there is no such thing as a “correct” grouping. The choice of what constitutes a group is completely subjective, it all depends on how far you choose to zoom in or out.

There absolutely are correct and incorrect ways of partitioning data into groups. Human genetics is no exception. https://en.wikipedia.org/wiki/Cluster_analysis https://en.wikipedia.org/wiki/Haplogroup

There are correct ways to group but there is no correct group size. Clustering algorithms inherently require some choice of sensitivity. There is no single correct choice for a given set of data, it all depends on what you’re trying to accomplish.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#35

Post by cyanydeez »

rahimnathwani wrote: post #29
cyanydeez wrote: post #24

Earlier quoted context omitted.

Paradox of intolerance answers your dilemma. Other answers: if placebos work, why would we restrict peoples choice and their dissemination? Because placebos rarely are innocent in often they become avoidant measures away from empirical science. Your view point is a local minimum in that you want to ignore societal impact. One of those impacts is using valid data to tout pseudoscientific methods. We know GIGO but the…

The paradox of intolerance suggests we should (if all else fails) suppress those whose seek to suppress others. How is that relevant? Is Lasker seeking to (let alone succeeding at) suppressing voices that disagree with his own? He might not engage with his critics (e.g. David Bessis) but AFAICT he's not doing anything to suppress them.

Surely this time race science wont be used to make social policies that directly boil down to racism by visual difference.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#36

Post by terminalshort »

cyanydeez wrote: post #35
rahimnathwani wrote: post #29

Earlier quoted context omitted.

The paradox of intolerance suggests we should (if all else fails) suppress those whose seek to suppress others. How is that relevant? Is Lasker seeking to (let alone succeeding at) suppressing voices that disagree with his own? He might not engage with his critics (e.g. David Bessis) but AFAICT he's not doing anything to suppress them.

Surely this time race science wont be used to make social policies that directly boil down to racism by visual difference.

So if I understand correctly, your argument is:

1. This research may prove that there are significant intelligence differences by race.

2. Public knowledge of this fact could lead to discrimination on an individual level based on group membership.

3. This is bad for society.

4. Therefore we should not conduct research down this path.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#38

Post by just_once »

terminalshort wrote: post #27
amanaplanacanal wrote: post #23

Earlier quoted context omitted.

The problem here is that "black" can mean anybody with dark skin from anywhere in the world. The sickle cell stuff is likely related to the fact that most "black" people in the US are descended from slaves that pretty much all came from the same small region in West Africa.

My point here is that this problem only seems to be brought up when the research has to do with intelligence. If you talk about genetic differences between "black," "white," or any other racial grouping on any other metric nobody ever brings it up as in my example above. So, while I acknowledge the fundamental weaknesses of the category, I have to conclude that the real objection here isn't the categories, but the to…

Most of the examples that you've used gain very little from added specificity. It's essentially linguistic laziness. That linguistic laziness is not identically consequential in all contexts.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#39

Post by rustyhancock »

rahimnathwani wrote: post #9

It seems weird for the government to collect data useful for research, but then to gate access based on the viewpoints of potential users. This restriction doesn't support the search for truth. Whatever people think of Lasker (Cremieux) and his views, isn't data being available to all interested parties the best way to find the truth? The article mentions this towards the very end: Adam Candeub, the top lawyer at the…

Scientists, are not quite like explorers, just venturing off in direction and see what they find.

They have interests, that align with funding, which aligns with actionable data. And they follow them.

So who would even do such research? The best you could find is some meaningless difference in mean scores that is likely swamped by environmental factors. And the fact that "race" is far fuzzier than our intuition leads us to think (it turns out we're very good at applying racial labels to people, that genetically are simply not so clear especially at the edges of categories).^*

So it's hard to fund research with no purpose, that scientists aren't particularly interested in conducting. Instead [GWAS](https://en.wikipedia.org/wiki/Genome-wide_association_study) studies answer more useful questions that are actionable (you can do a genetic test and calculate a risk in principle).

* I'll clarify this statement because it's a common misconception. There are clearly genetic differences between racial groups. But they are complex statsical genetics that are impossible to cleanly pin down without large levels of miscategorization. And GWAS studies are answering the question which genes are associated with cognition and disorder.

Re: Genetic data from over 20k U.S. children misused for 'race science'

#40

Post by rustyhancock »

The simple answer is that Race science is pointless because race is not a clean objective concept. We are fooled by our remarkable capability to assign racial labels into thinking it must be simple to cleanly deduce race from a generic test.

But it's obvious that races do have genetic differences.

The way scientists resolve this conflict is to bypass race entirely. And focus on genetics.

Genome wide association studies (https://en.wikipedia.org/wiki/Genome-wide_association_study) is the answer.

You create statistical models from large caches of genetic data and find association's. Then you can provide a risk score for an individual and if you want estimates based on ethnicity.

It's the genetic test that makes it useful like Duffy Null testing before starting people or African descent on Clozapine, HLA-B*1502 testing before starting Asians on Carbamazpine.

Where risks are lower you can do crude heuristics like calcium channel blockers before ace inhibitors for black people.

But just having people check a box in a questionnaire and doing their IQ is just pointless scientifically.

Post reply on HN