Humbuggery

This was originally going to be a reply to a comment on reddit here. But unfortunately it grew to 11000 or so characters, and Reddit wouldn’t have that. Surprisingly, this is the first time it’s happened to me, even though I have a penchant for incredibly longwinded comments. Enjoy this little TL;DR Internets.

Let’s start at the beginning.

It’s true, I’ve never taken a “real” IQ test. I have read _quite_ a bit about them though. I’ve cited some links which are representative of the materials I’ve read, there is more- and if you feel that those sources are somehow biased, I am very open to being presented with new evidence.

First, for those not aware, I just want to summarize the history and original purpose of the Stanford-Binet IQ Test and the Wechsler Adult Intelligence Scale, another important intelligence test which is based on an earlier test due to Wechsler-Bellevue[7]. Note that I’ve cited the Wikipedia entries for these ([2,7,8]) as being relatively representative, I’ve read other materia that. To be frank, I don’t want to bother finding, this is just a silly internet argument after all. One that does come immediately to mind is Stephen Jay Gould’s The Mismeasure of a Man.

The Stanford-Binet (SB) IQ test was the first intelligent test devised, it started with the psychologist Alfred Binet — after being commissioned by the French government — attempting to develop a method for identifying children with low intellectual ability or intellectual defect for placement in special schools[8]. Binet himself noted that this test should be used for the sole purpose of identifying such children[1] and was not to be used as “a general device for ranking all pupils according to mental worth”[1*]
Binet recognized that intelligence could not be describe by a single number, that theissue was fundamentally more complicated than that. The test (and associated Simon-Binet scale) he created was for the purpose of categorizing children.[1,2(Criticism,Binet)].

The latter test (the Weschler Adult Intelligence Scale/WAIS) was, as mentioned, derived from an earlier test[7]. Weschler defined intelligence as, “The global capacity of a person to act purposefully, to think rationally, and to deal effectively with his/her environment.”[7*]

A note of pedantics here, This is not what I would consider the colloquial definition of intelligence. Most associate intelligence with education, I want to make clear, they are, at least as far as the WAIS is concerned, different. Further, I would like to note that this is not a particularly precise definition- it encompasses all action, rational or irrational, as long as it is purposeful, but what do we mean by purposeful? Do we mean simply one whose actions are “under control” in a physiological sense? If so, then that would seem to imply that someone with a Motor-control disease, like Parkinson’s Disease, would be unintelligent- at least in that regard. If it means to that actions are well-thought out and rational, than isn’t the definition redundant? As the next criteria is the ability to think rationally? What does it mean to deal effectively with ones environment? What does it mean to deal effectively with anything? Again, I think the only logical conclusion is that the definition is equivalent to saying “Intelligence
is the ability of one to think rationally.” Which, I might note, is not a particularly bad definition, except that it is fairly narrow, and really attempting to test only logical ability. An important note for later.

The WAIS consists of 14 subtests and results in 3 scores. A verbal IQ, a performance IQ, and a composite, or “full-scale” IQ.[7]

Now let’s take a look at the subtests of the WAIS. It should be noted that, between the two tests, I find the WAIS to be better overall, but still lacking.

* Information subtest
This tests, for lack of a better term, social trivia. The example given in [7] is “Who is the president of Russia?” I argue that this is not a test of intelligence by the definition given, or more accurately by the constructed result — namely that intelligence is equivalent to the ability to be rational. This does not test my rational reasoning skills, my logical capabilities
or otherwise.

* Comprehension subtest

Similarly, this tests ones social “awareness” or “common sense”[6]. Again, I argue that this does not fit the definition, and is superfluous.

* Vocabulary subtest

A quick note- while I find that this too also fails to meet the definition- it is important to realize that the WAIS is an english-based test, non-english speaking minorities will naturally score lower here. Of course, one could simply translate the test, and we will assume that such would happen in the event of administration of such a test.

Now, I could go through every subtest, but it will be significantly quicker to simply state those I think do fit the definition. You may retort here, “This is a strawman, you’ve simply constructed an easily defeatable WAIS test.” But- I argue (obviously) that I haven’t. My argument is thus, the operational definition for the WAIS test is shoddy and redundant. I attempt to eliminate the redundancy, out shakes the parts of the test which follow only from the redundancy, and we are left with what remains.

Here is my list of “qualified” (in the sense that they contribute to a measurement of intelligence by its definition) subtests:

* Similarities, but only to an extent where the questions are fundamentally noncultural. Eg asking a someone who has never seen an apple or pear what the difference between the two is will obviously lead to a poor score. Not because the person is unintelligent, but rather because they have no cultural knowledge of _your_ culture, rather than having no knowledge of _any_ culture.

* Arithmetic, assuming it involves basic proof-esque questions. Eg “Find the error in this chain of reasoning” or “complete the argument.”

That’s it. Two out of the 14 tests really seem to fit the bill at all. So I’m forced to ask, How are either of these

a) Representative of the intelligence of the entire person?
b) How do either of these give any kind of implication on Vocabularial Intelligence?

Perhaps you can clear that up for me.

Before we get to the SB-IQ test, lets talk generally about some criticism of the _idea_ of an “Intelligence quotient”

First, we have to ask ourselves, “Is intelligence static?” That is, can you become smarter over time, or are some people naturally more intelligent and others not so. I’d argue that, depending on how you define intelligence, you can get any answer you like. Therein lies my greatest issue with IQ and intelligence psychometry in general, Either the definition they give is uselessly narrow- like the reduced WAIS definition above- or they are uselessly broad, and generally thereby redundant, like the original WAIS definition. If perhaps we were able to nail down a satisfying “Just right” definition, I would be more apt to accept these tests. But personally, I think the river is too wide to cross when it comes to intelligence.

Binet, as we have seen above and in [1], thought that his test and similarly intelligence as a whole was a non-fixed quantity. That is, your intelligence could fluctuate over time. Looking at the SB-IQ test itself lends why that might be true. Specifically the SBIQ Test covers (generalized slightly for space):

* Verbal Comprehension and Vocabulary
* Spatial / Visual Reasoning (Paperfolding problems, Mental Shape Rotation, etc.)
* Math/Arithmetic
* Memory
[from 9]

Now- let’s examine each. The first- Verbal skills- gets better as you get older. Certainly a 10yo child speaks with better vocabulary than a 5yo child, and similarly a 15yo better than a 10yo, and so on. Our vocabularies grow as the result of social interation. Similarly our Mathematical and Memory related abilities grow with age as well. We know from Psychology[10] that even the most basic kinds of memory
like Object-permanence don’t start until around 2yo. And similarly concrete reasoning skills only start around 7-8. It takes us till 11-12 to get to the point where abstract reasoning is possible[10] I would similarly argue that Spatial and Visual reasoning grows with age as well.

So- There you have it, that more or less sums up the research I’ve done modulo the time I’m willing to commit here. I just want to cover one other topic for another commenter. Specifically about Class/Income and IQ. Since these replies are fundamentally connected, I don’t want to separate them, so for the remainder of this post, I am addressing user:thaumaturgy, and not user:contrarian.

First of all, this is a very common issue, and has historically been a source of great debate (_viz_ [11]), I would argue that access to knowledge is a integral part of both of the most common IQ tests (WAIS and SBIQ, and WISC (in effect, WAIS for kids) too) and fundamentally controls the result. As you can see above, the better part of these tests boils down to education. I score highly on math tests because I have studied it intensely- I learned how to do speed arithmetic and learned how to speed-read and comprehend. Therefore I score highly on those sections of the test. Speaking in a general sense, the question boils down to- are these IQ tests testing how educated you are, or are they testing how many tricks you’ve learned to beat them. I would — as you may imagine — argue the latter. If a child is born to a lower-class family, does not have access to good education, does not have time to sit and study in the library, he will score — in general — lower on the SBIQ or WAIS. On the other hand, a child who was born to a middle-upper class family, had access to good education, access to the internet, access to a library and time to make use of these things will score better. IQ tests measure education, class influences heavily on education, vis a vis, class heavily influences IQ. If you feel that this is a poor argument, you are entitled. Perhaps I cannot see the fallacy I’m making, but it seems to me that the question that IQ purports to answer is “Are you intelligent” and a necessary followup question is “Is this nature or nurture”. My response is that IQ does not answer the question it purports two- but I still answer the latter as “both.” There is obviously going to be some element of nature involved in intelligence. Nature has a nasty habit of getting into everything- regardless of whether we think it should or not, but also I think it’s obvious that nurture, and specifically class-status, similarly influences your ability to be educated, and thereby your IQ scores.

* I lifted the quote directly from this article. I trust it can be independently sourced to
be from Binet.

[1] http://www.audiblox2000.com/dyslexia_dyslexic/dyslexia014.htm
[2] http://en.wikipedia.org/wiki/IQ
[3] http://www.lehigh.edu/~amsp/2007/12/gladwell-problems-with-iq-tests.html
[4] http://www.buzzle.com/editorials/3-30-2005-67806.asp
[5] http://en.wikipedia.org/wiki/Race_and_intelligence
[6] http://www.personality-and-aptitude-career-tests.com/about-iq-tests.html
[7] http://en.wikipedia.org/wiki/Wechsler_Adult_Intelligence_Scale
[8] http://en.wikipedia.org/wiki/Stanford-Binet_IQ_test
[9] http://www.answers.com/topic/stanford-binet-test
[10] The exact reference escapes me, but it’s in the Development section (Ch. 3) of _Invitation to Psychology_
by Wade and Tavris, 4th edition, 2008
[11] http://en.wikipedia.org/wiki/Race_and_intelligence