> Removing admission tests was a huge slap in the face to social mobility out of the middle class.
I agree, but arguably the benefits of the SAT with regard to class mobility is less than it was when it was introduced.
The heritability of IQ, combined with assortative mating, means that increasingly class divides are forming around differences in IQ. I think many people are rightly concerned that this will eventually result in simply a different entrenched class hierarchy.
I still think MIT should use the SAT, but the problem of maintaining class mobility is one we should continue to assess.
The heritability of wealth, combined with assortative mating, means that increasingly class divides are forming around differences in wealth. I think many people are rightly concerned that this will eventually result in simply a different entrenched class hierarchy.
That's how class divides have historically worked, yes. Previously those differences were also enforced through explicit class-based laws (the actually meaning of privilege).
The introduction of free markets, and later IQ testing, disrupted a great deal of the entrenched social class structure within western societies and allowed many people of rather humble backgrounds to achieve great things.
The concern I express is about whether IQ testing will continue provide the same level of social mobility in the future.
The research on IQ heritability is abundant. You’ll get death threats if you point out the research though.
It’s inconvenient and correlates with SAT performance. Hence the big push to get rid of the SAT to try to hide the disparities in general cognition within groups and between groups.
Maybe. I'm just more than a little skeptical given that GP has a scant post history, and spends his other posts railing on about CRT being marxism.
There's a lot of people pedaling this kind of IQ stuff to justify blatant racism, so you'll forgive me for wanting to see the data and not just letting statements like that fly by unchallenged.
You are right to be skeptical. Charles Murray is not the most reputable of scientists, and his book The Bell Curve is full of bad science to say the least, used to advocate for conservative (if not racist) policies.
A fairly quick summary is available on the Wikipadia page for this book (https://en.wikipedia.org/wiki/The_Bell_Curve#Reception) where you can dive deeper by looking into the reference (I particularly recommend S.J. Gould which addresses scientific racism in length in the book Mismeasure of Man). There is also a long video essay by Shaun (https://www.youtube.com/watch?v=UBc7qBS1Ujo) —if you have 3 hours to spare—which does an excellent job of summarizing the bad science and political agenda of The Bell Curve.
Making a big deal of the controversy around The Bell Curve and then going on to recommend The Mismeasure of Man is deeply ironic.
And why bring up the alleged racism when we are talking about class? The Bell Curve barely mentions race, and I'm not sure if Coming Apart does at all.
I'll watch the video at some point as it actually reviews The Bell Curve but I urge readers to seek out what Murray has himself said (both in his books and in interviews). The smear of him as being a racist and political hack is quite without merit.
It is true it is a bit unfair to claim that the policies Murray were racist without providing context. Some of the research he used to back up his case definitely were but the policies he advocated for were for sure conservative but only racist by association (which is why I put it in parenthesis). Conservative values in the 1990s America were really really connected to race. The whole debate about “inner city crime” and the notion of the “welfare queen” were largely a dog whistle. And conservative policies which advocated for “tough on crime” and welfare elimination were most often doing so against a largely racialized group.
I admit this is far fetched and maybe unfair. However I’m hardly the only one to accuse this book of being racist. The Wikipedia page is full of accusations, ranging from the aforementioned data from racist studies used unapologetically, to its advocates receiving funding from organizations associated with white-supremacy.
EDIT: I also want to address why it is totally fair to advocate Mismeasure of Man as a counterargument to The Bell Curve. There is a level of scientific rigor which The Bell Curve fails and Mismeasure doesn’t. For example the studies used in The Bell Curve were—to put it bluntly—bad. The Bell Curve failed to go through peer-review, it misused concepts, relied on unbacked assumptions, reached a conclusion unrelated to the premise, etc. You might find that some of these criticism applies to Mismesure, but hardly all, and not nearly on the same level. As an objective—albeit arbitrary—measure, just look at the ratio of the Wikipedia articles for each book which is devoted to criticism and controversies. Almost the entire article about The Bell Curve goes into some controversies. If both books were criticized and surrounded in controversies, one was definitely worse then the other.
Normally I'm glad to have a good debate about whether Charles Murray is racist, but the point I'm trying to make here is that whether he is or not is irrelevant.
Here we are talking about class differences. That's the primary subject of The Bell Curve (indeed he only devotes a scant few pages to the subject of race) and the only thing relevant to this discussion. Even if he were racist, them implication that he would have nothing important to say on matters of class is absurd.
> just look at the ratio of the Wikipedia articles for each book which is devoted to criticism and controversies.
Wikipedia is a great source finding out that things exist and their connections to other things, but there is no reason to believe it is objective with regard to politically charged issues.
That is not to say that the criticisms are wholly without merit, but few are actually technical critiques of the data or analysis.
Which is actually kind of sad, because while the technical critiques may be good (I haven't the time yet to check) the non-technical critiques I've studied are pretty poor.
> If both books were criticized and surrounded in controversies, one was definitely worse then the other.
The counter argument to that is that going against the shibboleths of our modern era is going to get you more flak, regardless of whether you are right or not.
This does not mean that Murray is correct because he is going against the grain and gets a lot of negative attention. Simply that all of the negative attention tells you a lot less than you might otherwise hope.
Regarding The Bell Curve, it being a poor academic work (i.e. technical critiques) is probably the main reason it is so heavily criticized. It used deeply flawed studies to back up its case, the conclusions did not logically follow the premise, it relied on dubious assumption with weak justifications, etc.
The fact that it then goes on to advocate for conservative policies based on such a weak case is why people conclude it is simply a racist work masquerading as academic. In an alternative world where the premise wasn’t of such poor scientific quality and the findings were sound, I bet people would still criticize it, but the overall reaction would be different. We would be trying to find a way to accommodate groups of lower IQ. Social scientists would try to identify barriers and advocate for their removal etc.
However given this work’s poor scientific quality, the whole notion that this group different a) exists, b) is significant, c), is innate d) is immutable, and e) largely inherits, should simply be ignored as false. Our current scientific understanding does not allow us to conclude this, or at the very least. This book is wrong in reaching these conclusions.
I appreciate the civil perspectives of you both, and would be interested in seeing you two discuss this in written format. I'd admit to having a relatively poor view of Gould, but this is not an impression formed from particularly close reading.
> Regarding The Bell Curve, it being a poor academic work (i.e. technical critiques) is probably the main reason it is so heavily criticized. It used deeply flawed studies to back up its case, the conclusions did not logically follow the premise, it relied on dubious assumption with weak justifications, etc.
The main reason Murray gets criticized is for his alleged racism and his politics. The attempts I've seen to discredit his actual arguments tend to be more self-righteous than serious.
I haven't read all of The Bell Curve (though I have read the sections that deal with race) but I have read other works of his, as well as criticisms of those works, and to my layman's eyes I find little credit can be given to his critics. Indeed, Murray often goes to great lengths to highlight to what extent each of his claims can be supported by facts, and does a great deal to acknowledge where there is disagreement in the scientific literature.
It may be that good critiques exist, but they are hard to find in the deluge of low quality hit-pieces.
> The fact that it then goes on to advocate for conservative policies based on such a weak case is why people conclude it is simply a racist work masquerading as academic
Frankly, this is why it's so hard to take his critics seriously. The only practical effect of the political Left's use of the label "racist" is to inhibit critical thought. We've spent much of this discussion talking about his alleged racism, when it still has nothing to do with the actual core thesis of The Bell Curve or my initial reasons for citing him.
> In an alternative world where the premise wasn’t of such poor scientific quality and the findings were sound, I bet people would still criticize it, but the overall reaction would be different. We would be trying to find a way to accommodate groups of lower IQ. Social scientists would try to identify barriers and advocate for their removal etc.
What do you think Murray's advocated policies entail?
Regardless, in a world in which people acknowledged group differences, there would still be genuine differences of opinion in how best to address those differences.
>However given this work’s poor scientific quality, the whole notion that this group different a) exists, b) is significant, c), is innate d) is immutable, and e) largely inherits, should simply be ignored as false. Our current scientific understanding does not allow us to conclude this, or at the very least. This book is wrong in reaching these conclusions.
That group differences exist is an incontrovertible fact. The extent to which those differences are driven by genetics is certainly up for debate, but Murray never claimed (and indeed, specifically argued against) that those differences were completely genetic.
The idea that genetics has no role to play in group differences is quite astounding, and attempts to defend that position tend to fall afoul of Lewontin's Fallacy[1].
> What do you think Murray's advocated policies entail?
He and Hernstein advocated for abandoning affirmative action, and divert funds away from supporting low IQ groups (e.g. welfare) into supporting high performance groups. This conclusion would always be debated among political lines as more left wing folks would argue against it in favor of accommodations, even in a world IQ was actual proven science.
> That group differences exist is an incontrovertible fact.
This is true, however the reason for this difference, and whether it is of societal significance, is up for debate.
I would argue that we should care more about SAT scores then IQ scores because SAT scores are actually used as an admission metric. These two correlate however I can just as well come up with a new metric which correlates with SAT and use that in my models in place of SAT, that doesn’t mean it is providing any additional info over SAT, nor that anyone should care about it.
> Finally, when we understand that the SAT is a reasonable measure of intelligence, we can use SAT scores as a proxy measure for time-consuming and sometimes unavailable traditional intelligence assessments, as dozens of researchers have been doing since 2004.
Although, I should say that Frey misses an important alternative interpretation of her conclusion here, which is to ask: “Why care about IQ if we can use SAT to measure the same thing?”
[/EDIT]
We do care about the group difference present in SAT scores, and we do care that high/low SAT scores seem to follow generations. However we see that as something to rectify. Nutrition, childhood led exposure, high noise levels all contribute to lower SAT scores, so we try to eliminate those. We could do the same for IQ scores, but why would you? If we found out that 50% of the variance of SAT scores were explained by genetics, who cares? The effect size is still tiny compared to other reasons why people score low on these test. Lets fix those.
Lewontin's Fallacy is a silly path to go down. You can draw taxonomy lines wherever you want. If you base that on a silly metric you get a silly taxonomy. Junk-in Junk-out still applies even if you subscribe to Lewontin's Fallacy. This is especially true in the study of human behavior.
Humans are a remarkably homogeneous species, we share almost all the same genes. Our experience varies way more then our genes. And alas, that is how most people split us into behavioral groups, e.g. by education, by socio-economic class, by occupation, by geographic region, etc. Group difference down these groups is way more interesting study then down gene make-up. And alas, this is where you find most research which actually contributes to policy change and further our understanding of the human mind.
EDIT 2: If you want to see for your self how poor the scientific quality of The Bell Curve is, I recommend you read the sources it cites. Of particular interests is the meta-alalyses by R. Lynn (1991) [Here is a summary https://youtu.be/UBc7qBS1Ujo?t=4256].
> He and Hernstein advocated for abandoning affirmative action, and divert funds away from supporting low IQ groups (e.g. welfare) into supporting high performance groups.
There are perfectly good reasons to support these policies that have nothing to do with being racist.
> This is true, however the reason for this difference, and whether it is of societal significance, is up for debate.
Agreed! But calling people with different conclusions on that question 'racist' is not debating. It is giving yourself an excuse not to have the debate, while doing nothing to convince the other side.
And Murray does not have a strong opinion about how much of group differences are attributable to genes. He only argues that they very likely play a role.
> I would argue that we should care more about SAT scores then IQ scores because SAT scores are actually used as an admission metric
Why do you think those are two separate things? There is not one single IQ test. Any test that is g-loaded is an IQ test. IQ is simply a way of representing a score on any test that measures intelligence. The SAT is an intelligence test, and you can represent scores on that test as a quotient.
> Lewontin's Fallacy is a silly path to go down.
It's a perfectly reasonable path considering you commit the same fallacy.
> Group difference down these groups is way more interesting study then down gene make-up.
There's no reason we can't study all of these things, including genetic differences in population.
> If you want to see for your self how poor the scientific quality of The Bell Curve is, I recommend you read the sources it cites. Of particular interests is the meta-alalyses by R. Lynn (1991) [Here is a summary...
I'll check out the video at some point, but it is very long so I'll have to find the time. A substantive critique would be appreciated.
Murray's most recent work on human diversity, Human Diversity, is also very much worth checking out, as it does far more to address issues, like race, that people seem to be so upset about.
If you want to hear from the perspective of someone who actual works in population genetics, you should checkout what Razib Khan has had to say on Murray's work.
The fact that genetic differences exist has nothing to say whether they matter. You can study those genetic differences all you want (like Razib Khan does[1]) but if you can’t show that these differences manifest in different behavior (which Murray tries to and has so far been unconvincing) then those differences don’t matter in the context of psychology. You can study it anyway, but you might just be looking for something akin to Russell’s Teapot[2]. While we don’t have conclusive evidence that genetics contribute a significant proportion of our behavior, then there is no reason to speculate around it as if it is a fact.
Now lets talk about Lewontin's Fallacy. First I’d like to note that Lewontin's Fallacy is not a logical fallacy. It is even debated whether we should call it a “fallacy” at all[3], especially when the metrics we use to group by are suspected to be biased. IQ is by no means a clean metric and we will never completely clear it of bias (even in a world were g was a proven fact, which it isn’t by a long shot). So if your metrics are biased and you measure a small group difference where inner-group difference is much larger then the inter-group difference, and the science is not conclusive in how exactly how this difference materialized in difference in behavior. Then applying Lewontin's Fallacy is silly at best (but much more likely it is disingenuous). If you insist that we should make a taxonomical distinction on such blurry lines your are just making an arbitrary choice. There are millions of equally blurry metrics which you could draw a million different lines. Why do you insist on IQ?
Finally lets talk about the difference between IQ and SAT. First of all Frey’s assertions that they are the same thing is actually a rather fringe on in the scientific community. They don’t correlate perfectly, they don’t even share a distribution, and they are designed to measure a different thing. The SAT authors repeatedly reject this:
> name change was meant "to correct the impression among some people that the SAT measures something that is innate and impervious to change regardless of effort or instruction.
1: Although from Razib Khan’s published works it looks like he is more interested in pumping the hype for consumer genomics and the exploring the cat genome then exploring human behavior through genetics, so he should not be taken as an authority—nor even an expert—in the intersection of genetics and psychology. I’m not aware of much research in this intersection, but if you want the intersection of biology and psychology, then the field of neuroscience has plenty. When I was studying the field over 10 years ago V. S. Ramachandran and Patricia Churchland were the big shots. Though this is largely off topic.
3: Although I’d like to say that a group difference with small effect size is probably still worth studying in the field of medicine. Especially in subfields where most of our studies are conducted on a single demographic of white abled bodied men.
I realized that I’ve shifted the goal post quite a bit here. From “The policies advocated by Hernstein and Murray can be considered racist” to “IQ is a scientifically uninteresting construct”. Or as you say—quite truthfully to be honest:
> But calling people with different conclusions on that question 'racist' is not debating. It is giving yourself an excuse not to have the debate, while doing nothing to convince the other side.
I did this a little on purpose (as you have noticed) because frankly the latter is a much more interesting debate (which I merge from a nibling thread I started where I made this claim).
So to address your concerns that I claim Hernstein and Murray are racist in their book The Bell Curve. First, I never called them racists directly. I claimed that their policies can be considered racist in the cultural context of 1990s american politics. I also claim that they blatantly and unapologetically use racist studies to back up their claims. So they are racists by association, and their work is racist as a result.
If they had made a convincing argument about the heritability of IQ and it’s importance in explaining behavior, and omitted citing blatant works of scientific racism to back up their case, I would be more hesitant in calling this work racist. I would just say that their conclusion is wrong. That we would need to make accommodations for people with lower IQ. We should keep affirmative action despite it being counter-productive simply because it is the right thing to do, etc.
However they did make a very unconvincing case that this g-factor explains the group difference in intelligence between groups. And they did cite racist studies to back up their case. And on top of that they used rhetoric to advocate for conservative policies which falls in line with what racists used at the time.
So to summarize my views on this less interesting debate. I consider Hernstein and Murray’s policy proposals in The Bell Curve to be racist because a) they use an unconvincing argument based on racist studies to advocate for set of policies, which b) were widely shared by other people at the time which were doing so because of their own racist beliefs. However you are of course free to disagree with my conclusion, although I hope you can see that these two premises for that conclusion have some reason.
> Maybe. I'm just more than a little skeptical given that GP has a scant post history, and spends his other posts railing on about CRT being marxism.
I do think the comments about CRT were a bit off-topic from this discussion.
However, I do agree that CRT is Marxist and I doubt the original formulators of CRT would object to that characterization. The people I've seen most vigorously object to that characterization have been classic Marxists who object that CRT rejects the materialism of Marxism. From just about every other perspective, there is a great deal of overlap.
> There's a lot of people pedaling this kind of IQ stuff to justify blatant racism, so you'll forgive me for wanting to see the data and not just letting statements like that fly by unchallenged.
Fair enough. However, most of the misinformation about IQ is coming from the skeptics of IQ[1].
I'm aware that IQ research is sometimes used by racists, but that's only possible because certain political interests have done their level best to make IQ as radioactive as possible. It makes it hard for anyone not already on the fringes of society to engage in the research.
To be frank IQ research is getting less and less relevant, and I suspect it has less to do with politics, and more to do with the fact that it has failed to further our understanding of human behavior and the human mind for the past 80 years.
Much more interesting research has come from other fields of psychology, including (in order of peak popularity) behaviorism, cognitive psychology, social psychology, neuro-psychology. Unlike IQ works from these fields have made predictions in unrelated fields and inspired lasting paradigm shifts inside psychology. For example social psychology has provided very useful constructs used in economy, behaviorism in criminology, neuro-psychology in medicine, cognitive psychology in computer science etc.
Scientifically IQ is a dead end. It only makes predictions inside the field of psychometrics, it correlates only with related constructs, the only place where it is used in an unrelated field is when racists are trying to excuse ludicrous statements.
Note: This is not entirely true. IQ was (is?) used to assess cognitive disability in individuals, but today we have a lot better tools for that (thanks to cognitive and neuro-psychology), so it’s usefulness has been surpassed by other fields.
> To be frank IQ research is getting less and less relevant, and I suspect it has less to do with politics, and more to do with the fact that it has failed to further our understanding of human behavior and the human mind for the past 80 years.
That's a fairly bold claim. Honest question: how did you come to that conclusion.
> Scientifically IQ is a dead end. It only makes predictions inside the field of psychometrics, it correlates only with related constructs, the only place where it is used in an unrelated field is when racists are trying to excuse ludicrous statements.
That's patently false. We already know it predicts academic ability (which is why MIT is reinstating the SAT requirement. It's an IQ test). And we also know it to be one of the best available[1] predictors of performance in cognitively complex work - like programming (Google's HR has publicly discussed this in the past).
[1]"best available" here doesn't necessarily mean great, just better than the competition. In domains like psychology that have a high degree of causal complexity, you rarely have any single factor that explains more than 50% of the variance in some subject - heck even 10% is pretty good.
> That's a fairly bold claim. Honest question: how did you come to that conclusion.
I didn’t conclude anything. This is merely a suspicion, I explained my reasoning below.
SAT being an IQ is a stretch. SAT is a scholastic amplitude, a far more narrowly defined construct. In fact IQ (according to psychometricion) should be immutable and somewhat inheritable, while SATs should reflect the work you put in during your school career and should be fair across demographics. IQ is also normally distributed while SAT has a negative skew (raw mean around 1050 with arithmetic mean of 1000).
You might have meant to say that SAT is correlated with IQ. While true, that is by design and is scientifically uninteresting on its own. SATs are far simpler and cheaper to administer. We have way more SAT data on the general population then IQ data. If your model includes scholastic amplitude but you measure it in IQ, you will not only have a less accurate model, but you will also make gathering data way more expensive and difficult.
> SAT being an IQ is a stretch. SAT is a scholastic amplitude, a far more narrowly defined construct.
SAT is an IQ test[1]. It was always an IQ test. There might be better (i.e. more g-loaded) tests out there, but the SAT still definitely works.
> You might have meant to say that SAT is correlated with IQ. While true, that is by design and is scientifically uninteresting on its own.
The SAT is correlated with g, which makes it an IQ test. You claimed that IQ doesn't correlate with anything other than performance on IQ tests. That is clearly incorrect with regard to academic performance. That you find this correlation 'uninteresting'is frankly beside the point.
> If your model includes scholastic amplitude but you measure it in IQ, you will not only have a less accurate model, but you will also make gathering data way more expensive and difficult.
There's not one 'IQ' test. Anything that correlates strongly with g is an IQ test. Yes, it also tests knowledge, but the ability to learn and recall facts is a heavily g-loaded activity.
Class divides have always been around IQ in every society.
Which is why CRT being pushed on children is so sick. They’re literally teaching children race Marxism. They’ve replaced the idea of class with race and now demand equity not between classes, but between races.
> Class divides have always been around IQ in every society.
Given by the positive effects of SAT testing on social mobility in the mid 20th century, I rather doubt that.
And to be honest, if we are going to have a class system one way or the other, IQ might be the better approach. But IQ is not wisdom, and I worry that the elites of such a society would be disposed to unprecedented levels of ego and hubris.
CRT is an outgrowth of critical _legal_ theory, which upon examination of the career of https://en.wikipedia.org/wiki/Roy_Bean has ample justification for existing.
I think he's saying that the whole society would be slaves. While I did not live under communism, I personally know some people who did and I think they would characterize that existence as a form of slavery.
I agree, but arguably the benefits of the SAT with regard to class mobility is less than it was when it was introduced.
The heritability of IQ, combined with assortative mating, means that increasingly class divides are forming around differences in IQ. I think many people are rightly concerned that this will eventually result in simply a different entrenched class hierarchy.
I still think MIT should use the SAT, but the problem of maintaining class mobility is one we should continue to assess.