Problems with IQ

8 Responses

  1. Matt says:

    Hi Charles,

    “ a ‘cognitive elite’ will and should ‘run a custodial state’ for an ‘underclass’ of people with lower intelligence that is ‘disproportionately black.’”

    This is the opposite of what Murray believes – he warns against the risk of the rise of a custodial state and doesn't like the idea of it; Malcolm Gladwell made a similar false attack on Murray (saying he wanted to put low IQ people on reservations, or whatever).

    I noticed a few weeks back that the New Yorker actually printed an apology/clarification…

    Don't rely on people with an agenda to read the book for you 🙂

    Or at least read Linda Gottfredson's public letter (in NYT or was it Wall St Journal?) – signed by 50 academics – about what academic research actually is on IQ, rather than what journalist and campaigners say it is. Indeed there's lot of great stuff on her academic website – discussing all the misrepresentations and misunderstanding about IQ research.

    Actually, I guess I ought to read the Bell Curve myself too…!

  2. Charles says:


    Thanks for the leads on some other reading. I checked out Gottfredson's letter and some of her other writing, and while I am on board with her desire to get the information and figure out what it truly means, regardless of our feelings about it, she seems a little too eager to get on the side of controversy.

    She seems more interested in pissing people off than finding the truth. And that doesn't sit well with me.

    For example, she supports statements that say IQ tests aren't culturally biased against American blacks. But that statement is only true in relation to school performance, because particular skills and acculturation are required to succeed in our school system. It does not apply to general intelligence.

    For example, the vocabulary used in traditional IQ tests can accurately be described as the vocabulary of "mainstream" American culture, which is mostly synonymous with "white" American culture. The classification as "white" is faulty, but it is a cultural division that is described, rather than prescribed, largely along racial lines. The result is that because of differences in cultural exposure, whites are more familiar with the vocabulary than blacks. This can hinder understanding of math and logic questions, as well as linguistic.

    When vocabulary is controlled for cultural differences, using terms that are known to whites and blacks, there is little difference in the scores.

    “Whites showed better comprehension of sayings, better ability to recognize similarities and better facility with analogies — when solutions required knowledge of words and concepts that were more likely to be known to whites than to blacks. But when these kinds of reasoning were tested with words and concepts known equally well to blacks and whites, there were no differences. Within each race, prior knowledge predicted learning and reasoning, but between the races it was prior knowledge only that differed.”

    Wikipedia entry "IQ"

    If this is shown to be true within a national distribution, it ought to be tested internationally as well. And until it is, the practice of using a global norm to compare national scores should be scrapped.

  3. Matt says:

    Hi Charles,

    I'm not sure what you mean by:

    "she seems a little too eager to get on the side of controversy".

    The controversy about the Bell Curve book was led by non-experts – she simply tried to outline what the accepted state of evidence in that subject is, as almost everyone was misrepresenting it. (As did the American Psychological Association, in their own report on this).

    So, if she you mean she was courting controversy, I don't think so.

    If you mean take one side in the controversy, then maybe yes. But if she's correct that the evidence all points to one side, then why not make that public. Admittedly many psychometricians are too scared to comment on IQ in public.

    Re cultural bias of IQ test – I'd got the impression that the whole discipline had spent decades removing any potential cultural bias. Indeed some (parts of?) IQ-related tests don't even include anything verbal.

    Whose research was it that found that:

    "When vocabulary is controlled for cultural differences, using terms that are known to whites and blacks, there is little difference in the scores"?

    I'd not heard before that you could just give a culture-fair IQ test and everyone comes out the same. Has that finding been published and held up under scrutiny?

    I certainly agree with you that if this is true, then IQ testing/norms are in big trouble.

    Have you perused Gottfredson's recent paper on the 'Logical fallacies used to dismiss the evidence on intelligence testing'.

    Does it provide any clarity on these issues?



  4. Charles says:

    It did seem to me that she is "courting controversy", but I could easily be mistaken. But if the evidence does point to one side, then it is the side to be on. But there's no harm in being careful with the hypotheses and testing methods while there is still debate.

    On the subject of cultural bias, I'll direct you to three links. First, an article in the New York Times from 2007 by Richard E. Nisbett, a professor of psychology at the University of Michigan, called "All Brains Are the Same Color". In the article he references other information, including the "Flynn Effect", and a study done by Drs. Joseph Fagan And Cynthia Holland that controlled for cultural bias in testing ("A Theory of Intelligence as Processing: Implications for Addressing Racial Differences in IQ"(pdf)).

    White and African American students were also tested for their understanding of sayings, analogies and similarities. When tested on sayings which required specific prior knowledge, White students averaged 65% correct, while African American students scored 48% correct. However, the results were quite different when equal opportunity to learn the information was available. For sayings that required only general knowledge, White students scored 72% while African American students scored 80% correct."

    The final link is to a "MENSA Workout". This is probably nothing like a legitimate test, but just a hokey example. In this 15 question test, there are 5 that require familiarity with specific vocabulary and idioms. Without prior exposure to those terms, an above average score is mostly out of the question.

    I read a bit of Gottfredson's paper, but I didn't see anything I hadn't read in her other papers. I've not encountered anyone who says the tests are unbiased address the issue of American dialects. As ridiculous as the concept of "ebonics" is, American subcultures speak languages as different as Chinese subcultures, and those differences correlate with subtle, but sometimes important cultural differences.

    As an experiment fill in the blank: "It's like grandma's ____________ pie."

    If you said "apple" you're wrong.

  5. Matt says:

    Hi Charles,

    Thanks for these great links – looking forward to reading them 🙂

    I think Nisbett was one of the authors in a great (special?) issue of an academic journal, which focused on IQ. It had lots of the leading pros and antis. (It may well have been this journal: Psychology, Public Policy and Law)

    In which case, I suspect that this was his bit:

    Nisbett, R.E. (2005). Heredity, environment, and race differences in IQ: A Commentary on Rushton and Jensen (2005). Psychology, Public Policy and Law, 11(2), 302-310.

    … but I just can't quite remember.

    I was impressed to find lots of key names all in one issue – and both 'sides'.


  6. Charles says:

    I'll look into that journal article soon. I've just picked up some economics reading, and I've still got to get through 4 chapters of Murray's latest. But I'll be back to it, I'm too intrigued with the concept of IQ and general intelligence to let it go for long.

  7. Matt says:

    Hi Charles,

    That Joseph Fagan material you mentioned certainly undermines the IQ proponents – it actually says that 'African American students' do better on a test about understanding sayings, if given an 'equal opportunity' to learn the sayings.

    This is quite a result, and I've never seen it mentioned in debates.

    That said, I never thought that IQ tests were meant to be about learning sayings or learning knowledge (by rote?). Fagan seems to say that the idea of IQ tests as being about processing or cognitive tasks is his pioneering insight.

    I always thought that's what they were. But what do I know, I'm not a psychologist.

    I also don't quite understand how African Americans are denied the 'equal opportunity' to learn about sayings in the first place. Can't anyone have a conversation, read a library book, or watch TV?



    PS I hope you'll post more about Murray's book. I hope to read it some time.

  8. Charles says:

    No one is denied equal opportunity to learn about sayings, but what a lot of people misunderstand is that the US has several regional dialects. And within those, there are subcultures. So the black subculture in urban Atlanta, will have use different idioms from a white subculture in suburban Denver.

Leave a Reply

Your email address will not be published. Required fields are marked *