More Thoughts On "IQ"
what the tests do and don't measure
As I’ve mentioned in previous posts on this topic, I don’t agree with the concept of “Intelligence Quotient”—it’s a measure that I find simultaneously vague, rigid, reductive, and extravagantly ambitious in its conclusion. All the hallmarks of bogosity. Few people realize that even Spearman added caveats to his claim that his tests provided a measure of “g” aka General Intelligence (defined as the common substrate for all abilities related to intellectual functioning…I know, arguably a tautology.)
I score very high on IQ tests. Or at least I did until my first encounter with the Raven’s Progressive Matrices test around 10 years ago, with my performance likely hampered by sleeping poorly the previous night. I also scored high on my SAT tests. I think that IQ tests that include a lot of verbal material, like the Weschler, measure something similar to the SAT: the level of competency with the skill sets related to scholastic achievement. And I think it’s accurate, as far as that goes. RPM is a very different test: people with mild-to-moderate verbally capable ASD have been known to score as much as 35 points higher on RPM, com[pared to tests like the Weschler “inventories”. Whereas my reaction to having the RPM test sprung on me was dubious, skeptical, and unmotivated. What’s weird is that both tests are alleged to measure the same capability—that mystical quintessence, “Spearman’s ‘g’”. Does that strike anyone else as kind of funny?
RPM is said to be a superior test of “IQ”, because of an emphasis on abstract visual diagrams rather than verbal questions that supposedly makes it free of culture bias. The possibility of other induced biases has since emerged. RPM views “abstract reasoning” as the baseline of intelligence. But—paradox, again—only for those who follow the rules! “Thinking outside the box”, so to speak, is discouraged.
My other big problem with the IQ concept is that it presumes to measure the ability limit of each individual tested. But it’s practically impossible to measure un-activated potential with one written test. I get that some people are smarter than others, but getting an idea of how intelligent someone actually might be requires a lot more personal interaction. (I mean, doesn’t it? Isn’t that basic?) IQ tests basically measure the ability to do schoolwork—or, in the case of RPM, the ability to use logical reasoning to follow patterns*, and that’s about it. It heelps to be well-rested before the test. Critiques of IQ and educational testing are also increasingly noting the influence of factors like hunger, lack of sleep, and lack of motivation on variability of test performance.
[* ability to follow the logic of RPM may provide a reliable measure of competence with computer “languages”. But anyone who tries to apply the logical prompts of the RPM to learning how to write and spell the English language is destined to run into multiple trainwrecks, often within the first paragraph. English is an Exceptional language, get it. You don’t get good by understanding the logic of it. You just have to get used to it. ]
<*edit>Speaking of abstract reasoning, I was just thinking that it plays a role in text-based verbal skills as well as mathematical skills and spatial diagram pattern recognition. Both skills are deductive. The difference is that numbers are denotative, and words are connotative. So math gets progressively more complex, sophisticated, and "difficult"--but what you see is what you get. Typically a set of clues and a mystery component to be completed to solve the mystery--an equation. The rules are clear, the numbers and the clauses and the operations are simply to be acknowledged and included to comprise the solution, which is easily reviewed, checkable.
But while text also requires abstraction and deduction skills for optimal reading comprehension, the process can be thrown off in the reader, subjectively, if they hit a "trigger word" ("Democrat", say.) Triggering is a language-based deduction that pulls in subjective content to influence the interpretation of the text. That obstacle can interfere with a reader drawing properly focused and germane inferences. It's also a property that allows a writer to manipulate language to serve an agenda--persuasive speech, rhetoric. Which includes propaganda, of course. So a high level of skill at abstract reasoning in math isn't necessarily transferable to the realm of textual verbal comprehension and reasoning, because the extra component of connotation makes for another aspect to be solved by accurately drawing out the proper set of inferences, and winnowing out the irrelevant tangential inferences. Clear close reading (and writing) partakes of the same step-by-step deductive reasoning as solving a math equation. But it's all done with Language, and all of the valences of connotation need to be tested, with critical assessment. One of the basis errors in that regard is to view words in isolation, like numbers, instead of contextually—regarding the way a word, phrase, or passage relates to the rest of the written text. Words aren’t to be read the way numbers are: as one label after another, to be summed in a linear process. Deductive inference also relies on language fluency and subset skills like jargon usage familiarity- empirical attunement built on practiced experience, induction. There’s no “natural knack” in reading, that confers superior ability. Reading and writing are more like learning to play a musical instrument. As with learning skill at playing an instrument, the key to obtaining a high level of reading skill is to achieve a basic level of competence that’s sufficient to want to do more of it. To want to continue practicing. But increasing skill at reading comprehension is more like developing the component of musical skills that consists of ear training—something to be learned by appreciation. If you love music, you’ll crave it and continue to get acquainted with more of it all the time. And your mind will be drawing connections to apply when playing your instrument, or in writing your own songs. In the same way, an ongoing practice at extensive reading acquaints you with improvements at drawing out inferences and meaning and symbolic resonance, and eventually influences the ability to improve at writing more clearly. Once reading feels like a true pleasure, like listening to a favorite selection of music, skills get transferred almost like osmosis. It’s a pity when someone has the basics to learn to love reading, and they instead consider it a chore. Their reading and writing skills are foredoomed to stalling out. My advice to anyone who wants to improve their verbal abstract reasoning skills is to find a topic you like—some non-computer, non-digital activity—and read some books about it.**
Note that oral verbality is even more ambiguous; that's why some verbal expressions, like legal contracts, are absolutely required to be recorded in writing in order maintain their power of enforcement, preferably in no uncertain terms. But oral verbality partakes of the additional component of social skills intelligence, which is largely about the sort of experience-based commonality of human sensitivity, expressiveness, and communicative affect that Justin Ruiz writes about here;
[*the words following were added 03/29/2025. ed.]
[**I don’t think there’s anything more important to learn about than the topics that Vaclav Smil covers in his writings. They tell readers about subject that the news media largely ignores: the reality of the planet we live on. Nature, natural resources, material culture and consumption. Books by Vaclav Smil ]
I know, I went off again. IQ is one of several hobbyhorse topics of mine. others include my talking up ranked choice voting; the practical, facts-based argument for a higher marginal tax rate for the top 10% of households, and especially the top 1% and 0.1%; my skepticism of claims for AI and unpiloted motor vehicles; and the urgent need for an overhaul of the substance criminalization laws. All of which you can read about on my Iconoclasms home page (still free!) Including more detailed articles on the topic of IQ, with link support included. https://substack.com/@adwjeditor/p-148984101