The Bell Curve

[Anonymous]

Does that make any sense?

Before I respond, I would like to point out that this sentence is as patronizing as "Does that help?"

Is there a reason I am the only responder in this thread to which, in every single response to me, you have included this kind of condescending ending?

Wrong. "Does this help?" is way more patronizing. In any case, the main reason is because you are one of the few people addressing the actual issues on a detailed level, and I sincerely wanted feedback about whether what I was saying made any sense or not. I like you and value your opinion and was hoping this would make you feel engaged in the conversation. I suppose when EA has made insinuations that I'm on the fast track to the Alt-Right for agreeing with Thomas Sowell about the quality of these chapters I could have said, "Screw you EA. Does that make sense?" Maybe I will next time.

As a reality check, here is what Thomas Sowell said about this book. Emphasis added:

The predictive validity and social implications of intelligence test results are carefully explored by Herrnstein and Murray in more than 500 pages of text with another 300 pages of appendices, footnotes, and an index. The Bell Curve is an education on the whole subject, including the evidence pro and con on a wide variety of controversial issues. Even where the authors clearly come down on one side of a given issue, they usually present the case for believing otherwise. In such candor, as well as in the clarity with which technical issues are discussed without needless jargon, this book is a model that others might well emulate.

It's fine if you disagree with him about how carefully the book is exploring these issues. But please be aware that being a flaming racist looking for rationalizations for your bigotry isn't the only reason to think that there is a lot of well-reasoned content in this book.

ETA: After adding the emphasis I decided to remove it. Didn't want to be patronizing, although I am still at a loss as to how somebody could think this is a negative review.

[_Analytics]

Re-stating the book's conclusion over and over in no way addresses the statistical issues brought up. Are you saying the book fails to address any statistical issues, or fails to adequately support their statistical conclusions? If the above summarizes the book as you see it, there is even less reason to take their conclusions at face value.

I don't understand the question. Can you give me a specific example of one of the statistical issues you are talking about?

[_EAllusion]

I vote you continue with the chapter summaries because I want to see what the book is about and I am really interested in how it either comes together or falls apart for you in the end. Most of us are taking issue in anticipation of part two and with some bias ourselves.

I think you have a legit issue in part 1 to settle with Lemmie, and I hope you can multitask and address that at the same time as reading, while ignoring most of the rest until you finish the book.

Part III if you are hanging on the racial issue. Part II is about the causal association of low IQ and socially undesirable behavior. Or, if you will, why poor people tend to be lazy and prone to crime.

[_EAllusion]

My fast track to the alt-right comment was a joke based on you citing a scientific racist popular with the alt-right to demonstrate how mainstream the Bell Curve is followed by accidentally implying that education doesn't improve performance on a aptitude test that measures academic knowledge thus throwing into doubt the value of public education as we understand it. That joke came after I gave both a detailed, substantive explanation of why Gottfredson's editorial didn't demonstrate what you claimed it did and why the raw assertion that little can be done to improve intelligence (as measured by testing) is dubious.

You keep strawmanning this notion that describing the Bell Curve as promoting racist pseudoscience means that any praise for any content therein amounts just being a hardcore racist looking for validation. While hardcore racists looking for validation are a considerable part of Bell Curve fandom, those aren't the same thing. It's entirely possible the arguments therein to have varying degrees of respectability which ultimately lead into outrageously dubious arguments on race and class that Sam Harris opts to promote. I think the book is a mixture of good science, bad science, and pseudoscience. I just don't think a good commentary on regression analysis or mainstream research on Spearman's g-factor makes up for the principle arguments of the book. Finally, you continue to misread Sowell's comments if you think he agrees with the main arguments of the book.

[_Lemmie]

Wrong. "Does this help?" is way more patronizing. In any case, the main reason is because you are one of the few people addressing the actual issues on a detailed level, and I sincerely wanted feedback about whether what I was saying made any sense or not. I like you and value your opinion and was hoping this would make you feel engaged in the conversation

Ah, well, what delicate flower of a girl can help but be flattered that a big strong gentleman would be so kind as to explain to her not only how her feelings of being patronized are wrong, but then also take very special care to make sure she feels engaged and not intimidated by all the big strong men who of course never need encouragement to feel engaged !!!!


I like you too, Analytics, but you have GOT to be kidding. What. The. “F”. You honestly think I damned need some special damned comments so that I'll feel damned engaged in the damned conversation? I mean, do you know me at ALL? I'm laughing my fuckin' ass off over here!!

But anyway, honestly, thanks for thinking of me, but I am the LAST woman who needs help to feel engaged in a conversation. Carry on, I'm going to go laugh for another hour or so!

Oh, don't worry about my stats questions, I've asked them enough times that if you feel like ever addressing any of them you can just answer an old post. Please, carry on with your reviews. I'm enjoying the thread more than I thought I ever would.

[_Always Changing]

I usually watch him doing his own thing. I just realized that even when the interests come together, there is still no communication.

[_Morley]

Regarding the accusation that they exaggerate the relationship, I'm at a loss. The accusation is made in a general way and I don't know how to address it.

Does that make any sense?

Upstream: Issues: Bell Curve: Thomas Sowell : Vol. 28, American Spectator, 02-01-1995, pp 32.

Perhaps the most intellectually troubling aspect of The Bell Curve is the authors' uncritical approach to statistical correlations. One of the first things taught in introductory statistics is that correlation is not causation. It is also one of the first things forgotten, and one of the most widely ignored facts in public policy research. The statistical term "multicollinearity," dealing with spurious correlations, appears only once in this massive book.

Multicollinearity refers to the fact that many variables are highly correlated with one another, so that it is very easy to believe that a certain result comes from variable A, when in fact it is due to variable Z, with which A happens to be correlated. In real life, innumerable factors go together. An example I liked to use in class when teaching economics involved a study showing that economists with only a bachelor's degree had higher incomes than economists with a master's degree and that these in turn had higher incomes than economists with Ph.D.'s. The implication that more education in economics leads to lower incomes would lead me to speculate as to how much money it was costing a student just to be enrolled in my course. In this case, when other variables were taken into account, these spurious correlations disappeared. In many other cases, however, variables such as cultural influences cannot even be quantified, much less have their effects tested statistically.

[_Gadianton]

I've read EA's criticism and Analytic's responses and Lemmie's recasts/sharpening of EA a few times and I'm a more than a little out of my depth here, but trying to understand what the dispute is.

Suppose I'm a hedge fund manager who trades on "volume" and "value", factors generally considered by that crowd to "modestly" predict stock price movement. Over thousands of trades and a little time, I make money. Now I quit the fund and become CEO of a big company. I read this bell curve book and bring my nerds over to devise a "trading formula" that hires and fires employees based on their 2nd grade IQ scores and salary data. Suppose the data is good data. I should be able to increase profits for the company (though not easily) over thousands of employees analyzed and given the shaft. In this sense the predictive power "matters for the group".

The dispute seems to me to be over the claim that the top 10% in wealth are the top 10% in intelligence. That would be like saying stock price movements are determined by volume and value -- with their "modest" correlations. when Analytics says, "The accusation is made in a general way and I don't know how to address it" I THINK the specific accusation is that modest correlations don't explain extreme stratification by intelligence?

In other words, if society were really that stratified, then epsilon wouldn't overshadow IQ for the individual. The epsilon for individual bench-pressers unlikely overshadows their tricep mass. Isn't the whole point of big data to figure out these subtle correlations and exploit them vs. finding totally obvious stuff?

The thing is, I can't totally pin Analytics down to saying what EA says he is saying, since there is this other way that little correlations add up in big data analysis to matter for a "group" consideration.

I'm reading Lemmie as cutting right to the heart and saying modest correlations don't predict individual or group success, the unidentified factors captured in epsilon do. It doesn't matter if a .004 correlation is much better than the next best factor of .0007.

okay...so all three of you feel free to take me to the woodshed in representing your views...

[_Always Changing]

okay...so all three of you feel free to take me to the woodshed in representing your views...

Good explanation, in terms closer to Analytics' field of expertise. Your background, I suppose, is economics.


https://www.samtiden.com/tbc/las_artikel.php?id=24

I was burned so bad by the pressure to practice School Psychology in conformity with LDS values and beliefs that I have avoided reading in the field for years.

[_EAllusion]

I think it helps to take a step back because I thought I was being sufficient vague in response to comments from Analytics I thought were vague. In the section that Analytics was commenting on, the authors try to build up a case that there has been substantial social stratification along the lines of IQ that is ongoing. I think this is what prompted questions from him that assume such stratification is substantial. Most of this case is circumstantial and based on just-so ideas that one should have a skeptical eye towards sans evidence. So, for example, the ability to perform well on mental tests is part of the filtering criteria for college entrance and having a college degree helps determine one's class status. In another example, there's been a growth in occupations since the turn of the 19th century that are more mentally demanding.

The main evidence they marshal in their favor is a body of studies that show a modest correlation between IQ and income and studies that show a relationship between individual job success and IQ. That just doesn't do enough to underwrite the stratification story because so many other potentially countervailing variables at work and the causal relationship isn't clear. Much is made of IQ being the best predictor, but the range of studies show a variety of factors that overlap into IQ in terms of correlative relationship. The relationship between income and the specific data set they use for most of the book exists on the low end of the range of studies they cite elsewhere and, if I'm not mistaken, would put it beneath the positive relationship between height and income as an example. Studies that try to cross-compare IQ with other specific factors tend to put IQ on top (though remember that IQ itself is pregnant with its own interrelated causes), but not in such a powerful way that mental ability can be thought of as having a determinant role on why class stratification looks the way it does.

All of this I think ends up exaggerating the picture. This is done rhetorically. Sticking to the evidence. IQ, performing well on mental tests, is modestly correlated with income.