African IQ in the US and Regression to the Mean

Chinedu writes:

But obviously today first and second generation Africans as a group are not only more successful than black Americans as a group, they’re more successful than many white ethnicities.

My understanding is that First generation Africans have IQ’s of 108, the highest or one of the highest of any ethnic group in the US. Given that Africa’s IQ in Africa is much lower, possibly ~70 IQ, it makes no sense that Africans would have a 70 IQ in Africa and their IQ would automagically jump to 108 in the US. IQ doesn’t work that way. It doesn’t skyrocket in one lifetime like that. In other words, obviously African immigrants are being highly selected for intelligence otherwise that 108 IQ would not even exist; instead they might have a 76 IQ.

Furthermore, the children of African immigrants are indeed crashing back down towards the mean African IQ, but they are not crashing all the way because the group is so high IQ to start with (108 IQ). So they are crashing from 108 down to 89 IQ. The children of African immigrants are ~20 IQ points less intelligent than their parents. However, an 89 IQ mean is very good for Blacks anywhere, so if US Africans maintain that 89 IQ, they will become one the most intelligent group of Blacks on Earth. Presently ,the smartest Blacks are Bahamains at ~93 IQ (although that figure is controversial) and 100 IQ for mulattos in Bermuda.

~90 IQ is better than ~85 IQ.

I do not think that their IQ’s will keep crashing down to 70 though, because regression to the mean doesn’t work that way. My mother told me that IQ’s tend to regress towards the mean, but they do not necessarily go all the way to the mean. In other words, if the White IQ is 100, parents with IQ’s of 140 may have kids with IQ’s of ~120, not 100. It won’t crash all the way down since it was so high to start with.

IQ’s cannot keep crashing down to the mean forever otherwise no group in the history of mankind would have ever experienced rising IQ’s and many groups have done just that. Regression towards the mean is very poorly understood by HBD’ers. You descend towards the mean, but not necessarily to the mean.

17 Comments

Filed under Africa, Blacks, Immigration, Race/Ethnicity, Regional, USA

17 responses to “African IQ in the US and Regression to the Mean

  1. Another William Playfair Web

    “My mother told me that IQ’s tend to regress towards the mean, but they do not necessarily go all the way to the mean. In other words, if the White IQ is 100, parents with IQ’s of 140 may have kids with IQ’s of ~120, not 100. It won’t crash all the way down since it was so high to start with.


    it will get infinitely close but never touch……kinda like this function, although actual rates of such do depend on hereditability.

  2. Another William Playfair Web

    Do you have a source on their IQ being 89?
    So, using the cranial capacity data, I once found an 0.60 SD difference, and with that 0.40 correlation 0.6(1/0.4)= 1.5 SDs below Whites, or on Flynn Norms, 103- 22.5= 80.5 or 80

    (108-89)/(108-80.5)=
    19/27.5
    or 0.69
    1-0.69= 0.31

    is heritability really that low?
    wow.

  3. Another William Playfair Web

    Interesting point:
    The brain drain could mean (depending on exact numbers) that African IQs should be higher with these immigrants still there- maybe a point or so?

    • EPGAH

      That’s true, but because we accept so many immigrants, they have the opportunity to come here. If you’re a smart person surrounded by the stupid and violent, you leave if you’re at all able to.

      It’s the basis of White Flight, and to a lesser degree, the cause of what is lamented as “Brain Drain”, whereby the ones who CAN contribute to society get away from a society that manifestly doesn’t want them or their contributions.

      I would LOVE to leave them there, but I don’t believe they’d make a positive difference, or indeed, ANY difference, I believe they would just get murdered by the losers they wallow with.

  4. TJF

    Rob and AWPW:

    Rob said:
    Regression towards the mean is very poorly understood by HBD’ers. You descend towards the mean, but not necessarily to the mean.

    I replied to Another William Playfair Web with my thoughts on regression to the mean:

    https://robertlindsay.wordpress.com/2016/07/13/photo-of-phil/#comment-265996

  5. Jm8

    The 89 figure (for the US) does not appear reliable in light of other data and achievement/socio-economic indications (even second generation Black Caribbeans in both the US and UK score above this, and they—at least in the UK—are substantially less selected and have an overall lower achievement profile than Africans) and the figure (I think) was claimed by Jason Richwine a while back ,supposedly—reportedly—based on only one moderately g-loaded subtest.,I think from his own report but not released.

    Second generation Nigerians (and Ghanains and Zimbabwean Shona) in Britain (and those brought by their families as children) do not regress below the first generation, but significantly (especially the Nigerians; both Igbo and Yoruba) outscore the White British general average (on the strongly g-loaded achievement tests), let alone the White working class.
    Second generation Africans (ethnicities/countries unspecified) in the UK generally exceed the first generation in education.

    All Africans combined (on the referenced achievement tests, also discussed in other posts); which includes (numerically very significant) groups;, less selected, unselected, and possibly even slightly negatively selected, like the Somali and Congolese refuges; score at about the White UK average-slightly below (often in the about same range as the Bangladeshis and Pakistanis).

    John Fuerst/Chuck tried to find data on 2nd generatio Africans in the US (trying to focus on West Africans/Nigerians. Apparently Black African/immigrant subgroups were not as clearly distinguishable in the US data he used, but he found the g-loaded achievement test) performance of the West African (likely heavily or at least substantially Nigerian) group was not below the general (White) average)

    http://www.unz.com/article/closing-the-black-white-iq-gap-debate-part-3/

    • The 89 figure (for the US) does not appear reliable in light of other data and achievement/socioeconomic indications (even second generation Black Caribbeans in both the US and UK score above this, and they — at least in the UK — are substantially less selected and have an overall lower achievement profile than Africans) and the figure (I think) was claimed by Jason Richwine a while back supposedly — reportedly — based on only one moderately g-loaded subtest. I think from his own report but not released.

      What are the IQ’s of second generation Black Caribbeans in the US and UK? I have heard that 2nd generation Black Caribbeans in the UK have IQ’s of 86.

      • Jm8

        The subjects is discussed in older posts including this one

        https://robertlindsay.wordpress.com/2015/11/04/uk-blacks-now-matching-uk-whites-on-achievement-tests/

        A few of my comments address/try to clarify) issues raised (in the comments) especially in one longish response comment where I link and quote a report by Steve Strand, with other linked sources.

        The performance of African subgroups (since in some reports they are not distinguished and they vary widely in achievement, degrees of selection, other ways) is in more detail in my link in the post before this.

        I commented there re: the tests
        “The Maths and English sections of the tests have about a .7 correlation to the g factor measured in iq, about the same as the SAT’s in the US. I believe different types of iq test have a similar correlation to one another…

        http://emilkirkegaard.dk/en/wp-content/uploads/Intelligence-and-educational-achievement.pdf

        The Maths and English sections of the tests have about a .7-.8 correlation to the g factor measured in iq,(To be exact; English correlation = .67 , Maths =.77)”

        And of course Bahamians Blacks are descended from slaves, and not from (often) positively selected (groups like) immigrants.

    • Another William Playfair Web

      Jm8-
      this all assuming additive heredity, where P= G + E, g is assigned a value between -1 and 1, and E is assumed to be the mean, hence taking P down towards the mean.

      If the IQs do not regress,as you say, it leaves the possibility that heredity is either 1.00 (which would really mean P= G , environment plays NO role, which of course is just silly)
      or
      Cross-reactions/norms of reaction, Genetics and environment are dependent on each other,
      i.e. someone’s genes will respond well to one environment but badly to another. If this was the case, it would really challenge the notion of “general cognitive ability” and IQ itself.

  6. Joe

    People conflate population IQ, mean regression, with sample IQ, mean regression. It’s actually in line, statistically, that you can create new means based on your sample population. We can understand this as “lineage IQ” or “ancestral IQ.”

    It just so happens that the variation of sample means produce a particular average. It’s not so that a high IQ sample will regress to the population mean – in fact, that’s a statistical fallacy of division. It is so that a high IQ sample population will regress to their ancestral mean.

    • Joe

      I think I might program a new simulation for this. I did one months ago, but it was random, and the algorithms are inefficient.

    • Another William Playfair Web

      yes, it’s mainly the population as a whole, and of course, hence the most likely value the offspring will be, (this is what Misdreavus failed to understand on “peepee’s blog).

    • Another William Playfair Web

      “People conflate population IQ, mean regression, with sample IQ, mean regression. It’s actually in line, statistically, that you can create new means based on your sample population. We can understand this as “lineage IQ” or “ancestral IQ.” “

      Joe, you are essentially saying it’s isolated from the (racial mean), if so,
      then what does it regress to?

      this is assuming that all genes are roughly the same, inherited from parents, and Environment defaults to it’s average, which has caused the effect to give the global IQ average of the race.

      • Joe

        “Joe, you are essentially saying it’s isolated from the (racial mean), if so,
        then what does it regress to?”

        It regresses to the sample mean: a racial mean is only an aggregate of sample means. A population mean is the trend, which can be used to determine the probability of sections of the population, not of a particular sample.

  7. Pingback: Regression to the Mean Means Regression to the Group Mean, Not to the Racial Mean | Beyond Highbrow - Robert Lindsay

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s