I'm surprised that the article talks about "specialization" without mentioning that many modern IQ tests have specialized subscores... but then again, this may

I'm surprised that the article talks about "specialization" without mentioning that many modern IQ tests have specialized subscores... but then again, this may not be common knowledge. Back in the day I used to compile patient evaluation reports for a psychologist who specialized in Diagnosing Patients With Stuff. The point of the diagnoses (and the reports) was generally to help kids and adult students get educational accomodations: for example, if a kid's report proves that they struggle with written instructions due to dyslexia, they'd be able to request a verbal/auditory alternative for relevant assignments. The psychologist would administer LOTS of tests and interviews for each patient, usually over the course of a few days, and those tests usually included one of the Wechsler IQ tests: either the WAIS-IV/V for adults or the WISC-V for kids. Both of these tests include an "overall" score called full-scale IQ (FSIQ), but this wasn't the focus of patient reports—the different subtests were. Also noteworthy that FSIQ is just the average of the scores on those subtests, which measure things like verbal comprehension, working memory, and processing speed (and sub-sub-tests thereof). These are useful in a clinical context because they let you identify conditions that are marked by discrepancies in different areas: for example, that kid with dyslexia probably struggled with processing speed but did fine with verbal comprehension. So a patient's overall score was a lot less relevant than any subscores that seemed like outlier compared to other subscores. (Today's Wechsler tests are aware of the Flynn effect and supposedly get recalibrated to account for it, but I can't speak to that personally.)