I waffle on how useful "IQ" is as a measurement. For example, I marked off twenty minutes on my PC clock and took this Mensa "lick the thumb, is it even worth time trying the official examination?" pre-test, compliments of the Luxembourg chapter. I got 24 correct out of 33.

"Congratulations", it told me, "you're not an idiot; you should try the real thing, you'd probably pass!" But I didn't feel very smart as I took it, even as the test encouraged me to go spend 40 EUR getting officially examined; here is why:

The "ideal" way to measure intelligence would be to separate knowledge from raw cognitive processing power. But IQ tests don't even remotely do that. They claim to, but they don't.

You can see this contradiction here, where these types of tests are described as not evaluating "general culture"-- so, knowledge-based things that only some people would know-- but rather "logical and mathematical skills", such as "the speed of information processing, analysis, and synthesis", which is presumed to be knowledge-agnostic.

And yet, when I took the online pre-test, I took one look at the sequence question that started with "65536", and filled in the final box without even having to think about it. Is it because I'm a genius? No: it's because I'm a full-time computer programmer! Powers of two are second nature.

Similarly, I prefer analog clocks to digital ones, and have lots of the former in my house. Several of the questions on the test featured patterns that went either clockwise or counter-clockwise. I picked up on those immediately, because my brain is conditioned by the constant knowledge exercise of how to read an analog clock.

It's sort of like this guy, who can play ragtime music pieces having never even seen them before. Is he a genius? No: he's a ragtime pianist! He has years and years of practice in recognizing common ragtime sequences, chords, timings, and so on. When he looks at the sheet music, he doesn't just see notes, similar to how when I look at powers of two, I don't just see numbers.

I was a bit stumped on a few of the Mensa pre-test's questions, but it was obvious to me that it wasn't because I was stupid-- it was because it was based on some pattern that I wasn't familiar with. Five seconds of someone explaining the pattern, and boom, I could go from a 125 IQ to a 140 according to one of these tests, because it would "open the door" for a littany of additional question types!

In that example, did my brain suddenly become better at "information processing and synthesis"? Nope-- but I acquired a new piece of knowledge, which caused me to get more questions right, which caused the computer algorithm to generate a higher "IQ" score.

To be fair, "IQ" test makers have their hearts in the right place. And I'm sure you could learn something about rough cognitive abilities by controlling for as many variables-- such as upbringing, culture, type of schooling, and so on-- as possible, especially if two scores are extremely disparate; it would be hard to explain away an 80 score versus a 140 based on knowledge alone.

But, I think "IQ" is trying to solve an insolvable problem. It's a fool's errand. And that's not even getting to how absurdly narrow its scope is; some people are bad at visual patterns, but can deconstruct verbal language statements with one ear closed, just to name one example.