The book Measured Lied: The Bell Curve Examined addresses Herrnstein and Murray’s book chapter by chapter, and rips it apart quite well.
Let me ask you a question, december: Suppose the findings of The Bell Curve were and are 100 percent accurate–what would you consider to be the prescriptive implications? What should it mean for public policy, etc.?
TBC takes certain positions on a long-running debate. In particular, the authors’ views are
[ul][li]General intelligence (g) exists. g has a substantial genetic component.[/ul]The book itself makes some policy recommendations. The one I recall revolves around the fact that there are many dumb people. Therefore, the authors recommend that public policy aim more at simplification and clarity. This echos the theme of The Death of common Sense. It’s a principle I deeply believe in. It’s something I have worked to promote within organizations I have worked for. One reason why I am about to leave my job is that my current employer doesn’t operate on that philosophy.[/li]
In terms of private policy, I have had good luck hiring smart people. I do more-or-less believe that there exists something like a “general intelligence.” I seek to hire people with that attribute. At one time my wife was a consultant at Stanford Research Institute. Her boss Chuck liked to boast that he was the dumbest person in his department. That is, Chuck would only hire people smarter than himself.
I don’t have much background in these issues, but I do find them quite interesting. How about these issues:
Almost all of TBC’s data, as I understand it, was based on the National Longitudinal Study of Youth (NLSY). The mean IQ was only 94, which is (fortunately) not representative of the United States population, and the standard deviation was not 15 or 16.
M&H address a white-only study in order to avoid a “complication based on race.” In their section on “Cognitive Classes and Social Behavior,” M&H describe the relationship between IQ and poverty, dropout rates, unemployment, crime, etc. I believe that this invites the readers to judge that these same relationships hold true between ethnic and racial groups, not merely among them.
M&H do not question the accuracy or precision of this study, which leaves their work open to criticism.
Furthermore, M&H appear to be of mixed mind about the Flynn Effect (the general trend of IQs going up). As I recall it they agree with this interpretation of the data in the beginning of the book, but then try to dismiss it later in an attempt to demonstrate that IQs are in fact going down among people in certain groups. I never really understood this inconsistency.
I do not know a whole lot about statistics, but I do know a little about regression coefficients and r-square. Has anyone tried to check M&H’s math?
Murray says himself that they “guessed” it was “about 40%”. To my knowledge, they offer no substantive proof one way or another as to the accuracy of this claim.
I cannot fault you too much for making the underlined statement, Maeglin, because it was widely repeated by critics. However, the critics never had a cite, because TBC never said this. In fact, the book denied this several times over. It made the point that one cannot compare the genetic component of white vs. black IQs, because of environmental differences.
M&H’s point was to try to demonstrate that g exists. This has been a controversial question for 100 years. It’s not really settled today. Many experts have contended that there are many types of “intelligence,” rather than a single one with broad impact. By showing a relationship between IQ and all those areas, M&H claim to show how meaningful general intelligence is. As you noted, they used a single race in this section precisely in order to focus on the meaningfulness of IQ, because they wanted to avoid adressing racial differences. It’s unfortunate that some critics refused to take them at their word.
I was fascinated by the book’s coverage of early intelligence research and debates. In particular, I learned that a number of commonly-used non-parametric statistical tests were devised for the purpose of IQ studies, including the Spearman rank-order corellation coefficient, which at one time was used in the process of calcuating workers compensation insurance rates in California.
Maglin,
I read someplace where the Flynn Effect might be the result of our moving from a relatively simple to a more complex world. The fact that our environment has become increasing more complex over the last 100 years and is increasingly bringing the whole society into that mix. I’m not sure whether I saw this in the Bell Curve or someplace else but the premise was that since IQ is determined by both heredity and environment - IQ has seen the “Flynn Effect” increase.
Monozygotic twin studies indicate a positive correlation at about .90 This is when raised together. When raised apart I think it dropped to around .75
This is all from memory – I’ll look it up later and retract if wrong.
I take your point, december, but my bigger issue is that I suspect the correlation between IQ and socioeconomic status (SES, I think some people call it) in the first place. Hence using a survey that tests substandard and environmentally underpriviliged youth rather overdetermines their results.
They admit that they don’t have a clue what the proportion between genetics and environment that seems to boundary intelligence, they merely posit that one exists. I don’t disagree with them substantially on these issues.
“Simplification and clarity”? The authors put forth (the old refrain that Murray had championed for the prior twenty years wonderful how their “science” just happened to support Murray’s political agenda) that we should, as a society, abandon all efforts to improve the environment of people who tested “dumb” and, in a plan somewhat analogous to the Japanese model, simply divide people into “smart and dumb” as they enter school, while looking for “dumb” jobs to get them when they come out the other end (having never been given the opportunity to leave the “track” they were assigned as four-year-olds).
I suppose that they might call that “simplification and clarity”.
Of course, since the overwhelming majority of their IQ conclusions were based on the corrupteds use of the Armed Forces Qualifying Test (which was never designed to measure “IQ”) with the claim that they had found a way to pretend it measured IQ, their science is more than suspect. Given that they made their science match their previously established philosophy, I don’t find their philosophy to be very honest, either.
Here is a summary of the APA findings noted in my link, above:
So, december seems to be accurate when he notes that Murray and Herrnstein got a lot right when they were quoting other people’s literature, but it seems that the professionals in the field found that their own meager attempts were just bad science. That they then published their “findings” as science while hiding from peer review, indicates that they were, indeed, dishonest.
An interesting article here that I’d love to see commented on by some of our worthies on the subject
http://www.brainmachines.com/body_wolf.html
Of particular note is the “IQ Cap” Which can accurately predict an IQ score on a standardized test simply by attaching a few electrodes and staring at a pin.
Also, an interesting essay on the politics of this kind of science.
Please let me know what you think.
This sounds like a parody, particularly the part about never giving an opportuniny to leve the track. Can you cite what the book actually says?
We are all aware that the book was highly controvesial and reaction to it was highly politicized. Some criticisms were unfair – just a political smear, e.g., the Scientific American book review. Other criticisms were made in good faith.
We can quote purple prose from critics and supporters, but I’d rather debate specific points if anyone has any.
I’m still waiting on someone’s response to my remarks on the NLSY. That not scientific enough?
I’ll try to take a look when I get home, if I have time tonight.
Scylla -
Thanks for the link. I think some of the other points made in the article were even more interesting than the brainwave / IQ connection. The social aspects of ‘theorys’ like this one that force people off their mark has even been reflected in some of the posts here — and may even be of more relevance in this debate than the subject matter itself. The article dealt with taboos that if mentioned in ‘political’ society will bring a predictable response despite the science involved. We seem to have a solid history in that respect. Yet - it remains a serious matter since these sort of response aren’t based on science yet they will prevent it if one isn’t careful. We all have to recall that Nature doesn’t bend herself to our ideas about right and wrong – and as Will Durant said, “The Sun isn’t there to light our cigars.”
On another note - I’d be interested in hearing ANY responses to not only Scylla’s post, but also to the twin studies which have been done ---- without resort to ad hominem which adds nothing and only clouds the valid points one might make.
I have, in general, tried to avoid posting when someone has already said what I was going to, unless it was , for instance, a poll in which a “me, too!” response actually contributes information. Otherwise, I’d simply be adding to clutter on the Board and taking up too much of my own time (I spend more than enough on the SDMB as it is!). In the present case there have been more than enough responses to december, and many have been much more thorough and better-written that anything I can have produced.
On the other hand, december addressed his challenge specifically at me – why, I do not know, as others had written in before – and that seems to require a response. I can’t let others on this Board fight my battles for me.
That said, I recognize my limitations. It’s been a few years since I read The Bell Curve and The Bell Curve Debate and other responses, and I’d have to bone up before I could give a useful reply. And when I did, this not being my specialty, I’d be parroting what those sources said.
My answer, then, is to direct december to the book I cited above, The Bell Curve Debate, as well as the others people have cited. My other alternative is to dig into the subject myself, and I really haven’t got the time. If you think I abdicate any right to respond to this or another thread because I fall back on these experts, then all I can say is that I doubt if you have personally researched or checked your sources completely, either. I am familiar with the writings of the contributors to the cited volume, and trust their work. I even know one of the authors personally, and can vouch for him.
I have, in general, tried to avoid posting when someone has already said what I was going to, unless it was , for instance, a poll in which a “me, too!” response actually contributes information. Otherwise, I’d simply be adding to clutter on the Board and taking up too much of my own time (I spend more than enough on the SDMB as it is!). In the present case there have been more than enough responses to december, and many have been much more thorough and better-written that anything I can have produced.
On the other hand, december addressed his challenge specifically at me – why, I do not know, as others had written in before – and that seems to require a response. I can’t let others on this Board fight my battles for me.
That said, I recognize my limitations. It’s been a few years since I read The Bell Curve and The Bell Curve Debate and other responses, and I’d have to bone up before I could give a useful reply. And when I did, this not being my specialty, I’d be parroting what those sources said.
My answer, then, is to direct december to the book I cited above, The Bell Curve Debate, as well as the others people have cited. My other alternative is to dig into the subject myself, and I really haven’t got the time. If you think I abdicate any right to respond to this or another thread because I fall back on these experts, then all I can say is that I doubt if you have personally researched or checked your sources completely, either. I am familiar with the writings of the contributors to the cited volume, and trust their work. I even know one of the authors personally, and can vouch for him.
Maeglin - They estimate that IQ is 60 percent heritable - with a range from 40 to 80 percent. This is based on the studies of twin, sibling and unrelated subjects - both raised together and apart. These numbers, as I understand, are typical – not unique to Herrnstein / Murray. Others have published similar results. So, it’s more a factor of Herrnstein / Murray publishing outside of the scientific commuunity that seems to bring most of the criticism –
First, if we want to discuss this topic, ** scientifically ** as opposed to a bunch of uninformed ideological blithering on, one has to go get the primary literature. Your half-recalled twin study is not up to snuff. They are also fairly irrelevant to the issue of the Bell Curve given the Bell Curve goes well beyond the general observation that one’s genetic background provides an underlying template for what eventually is expressed.
Why the first? Because in order to have some substantive comment one has to know what the actual research designs were, get a feel for how they addressed problems etc. These are non-trivial issues when dealing with a complex subject such as this. I would do so myself, but frankly I tire – as in the case of the race and genetics exchanges – of being the only person to provide a range of primary literature and comments. And I lack the time right now.
As to “ad hominem” attacks: december has never shown to date any desire or capacity to actually engage arguments. I feel confident that this is not an empty attack on another poster but an actual critique of his argument. See supra where he simply pretends (and this is a constant pattern which many, many other posters have called him on, conservative and liberal, those that agree with him and those that do not) that nothing contradicts his ad hoc assertions. In point of fact, for example, the Slate article gives several substantive methodological critiques of the work. As does Cal’s citation. If december were honestly interested or capable (or both) in engaging the debate, he would address the substance. He has not and he will not, he will only assert, backpeddle and ask others to do his work for him.
As to the issue of heritability, therein lies one of the problems: heritability does not ipso facto translate as genetically determined. This has been addressed in past Bell Curve debates here and elsewhere. Far too much of this is entire discussion begins with such naïve misunderstanding and then moves from there.
As for the characterization of much of the criticism coming because they published outside of the scientific community this is pure nonsense, and I refer you to Tomndeb’s helpful summary.
Described that way, it simply sounds as though people were irritated that they “went outside.” That is not the problem.
The problem is that they “went outside” the scientific community and published the book (to great hoopla) for the purpose of shaping public debate when their science had errors in it that could have been found and corrected had they gone through peer review. (Of course, correcting their errors would have destroyed much of their political intent, which was why they avoided the review.)
In other words, This is not an ad hominem attack. december really is an asshole. :rolleyes:
However, Collounsbury, you are correct that I am not passionate about this debate. I will look at my copy of TBC tonight and try to respond to Maeglin’s point. However, my level of interest is not high enough to sustain the debate on my own. If you or another poster wants to initiate discussion of specific aspects, I’m willing to keep the debate going.
BTW you didn’t respond on the use of the Spearman Rank Correlation Coefficient and other non-parametric tests. Doesn’t you nonpareil background lead you to have a position on those techniques? 