Michael Eriksson's Blog

A Swede in Germany

IQ myths

with 4 comments

A great annoyance when IQw is on the table is that many of the PC persuasion make cocksure, yet absolutely incorrect, statements about it—often “proving” the opposite of reality in the process.

For instance, on a recently discussed blog entrye (where I still cannot comment) someone today made the following two claims as a means to dismiss The Bell Curvew:

  1. IQ is not given by nature, which is proved by how easy it is to train it.

    In reality, IQ has comparatively little yield (less than e.g. a time for the 100 m dash), even having very strong correlations between results at e.g. ages 10 and 30. Even if training has an effect, the thing most likely to change is not actually the IQ proper (or, in its core, g), but the test taking ability. (A reason why some prefer to only give tests to pristine subjects. Note that the proportion of people who are repeated test takers is limited and the distortion of overall numbers, in particular between groups, is small.) Doing this might give better room for bragging, but changes neither reality nor the value of IQ tests.

  2. IQ is unlikely to be inherited, which is proved by unspecified claims about adopted Third-World children being closer to their adopted than native countries in terms of IQ.

    In reality, IQ is very strongly hereditary, which has been shown again and again by a great number of studies—including those involving twins placed into foster homes of different SES.

    As for the specific claim, the lack of source makes it hard to make a definite statement. (But I do vaguely recall having heard similar claims on other occasions.) However, the fact that Third-World children were chosen would make the applicability in general low: The Third-World has problems with e.g. nutrition that make the impact of environmental factors far larger than in the First World—and it is quite possible that merely removing these detrimental factors could account for more than half the gap. This, however, has no impact on the observations made in the First World.

    (Generally speaking, the relative influence of nature and nurture on a group will depend not only on their “absolute” influence, but on how much variation is present. For a simplified example consider the influence of x on f(x, y) = x + y when x varies between 0 and 10 and y between 0 and 1 resp. 0 and 100.)

Other popular incorrect claims include:

  1. Scientists like Stephen Jay Gouldw have disproved IQ.

    They have not: Experts see great value in IQ and Gould has been criticized for making unqualified claims, referring to a state of research decades behind even his time (and note that his The Mismeasure of Manw is another thirty years old by now…), misrepresenting scientific consensus, and building strawmen—with a more than good chance even of ideological bias on Gould’s part.

  2. IQ only determines how good you are at taking IQ tests.

    In fact, IQ shows a high degree of correlation with a variety of tasks. There is even a correlation between IQ and speed of reaction. It is true that IQ is a very imperfect (but far from worthless!) predictor for individuals. However, for groups it is very useful. Further, even for individuals it can make statements about e.g. what work positions are at all possible.

  3. IQ is “culturally loaded”, biased against non-White, non-Christian, non-European men, or similar.

    Decades ago, this was to some degree true. Since then, great efforts have been made to investigate and eliminate such problems. One of the purest tests, Raven’s progressive matricesw, shows the same general group differences as have so often been ridiculed as caused by cultural bias. Indeed, cultural bias can often reduce a group difference: By reducing the g loading and making knowledge and experience more important, a smaller difference in these areas will mask the difference in g.

Finally, there is a claim that is true, but often used in a misleading manner:

IQ and intelligence are different things and IQ does not measure intelligence

IQ is indeed only a proxy for intelligence, even increasingly a proxy for g. Notably, it can be argued that removing “culturally loaded” questions (e.g. relating to word knowledge) has made it a lesser proxy for intelligence… To some approximation, it can be said that IQ measures the inborn part of intelligence—which makes it highly valuable and allows it to (approximately) fulfill the demands that are put on it today. In an earlier entry, I compare basket ability and height with success in life and IQ, noting that it would be equally foolish to dismiss IQ for success as to dismiss height for basketball—it would be several degrees more foolish to dismiss IQ when talking about intelligence.

For a decent overview with many further sources, I recommend the original link to Wikipedia. This page is not perfect, often being altered by PC zealots, but the facts usually shine through.

Finally, I would like to throw in a recent Dilbert stripe that not only matches most uses of the word “racism” I have ever encountered on the Internet, but which is particularly apt for discussions around IQ, “The Bell Curve”, and similar topics.


Written by michaeleriksson

June 14, 2011 at 11:14 pm

4 Responses

Subscribe to comments with RSS.

  1. This topic is often difficult to discuss publicly and makes some people upset/uncomfortable. I’m sure you know why. (I agree with your disagreements.) Thanks for commenting on my blog … hope you stop by again!


    June 19, 2011 at 1:51 am

  2. I’d hazard a guess that your IQ is higher than mine. It is said that knowing more than one language actually makes you more intelligent. It sounds quite plausible.

    Bob Robson

    June 19, 2011 at 4:21 am

  3. Just gve a definition of intelligence and prove that it can be measured. —- If neither of this can be done, I guess we will have to manage with the situation as it implicitly is described by you, but perhap with a little bit more scepticism as to their use in different sortings and judgings of people . After all : we do need every “genius” that we can find.


    June 23, 2011 at 9:04 pm

    • I doubt that a consensus definition of sufficient usability can be found. By analogy, we can define how tall or heavy a person is, but stating how large he is a different matter—and intelligence is even harder to pin down. To make matters worse, not all definitions of “genius” require and imply intelligence (let alone IQ).


      June 23, 2011 at 9:53 pm

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