I’ve often questioned my intelligence. And I believe there’s something to IQ tests, although I’m not 100% sure they’re perfectly reflective of the academic abilities they’re supposed to measure. Nevertheless, I feel IQ tests provide a much more faithful measure of someone’s intelligence than people. Peoples’ opinions of your intelligence, I believe, are usually reflective of their own interests and they tend to be confused about the differences between intelligence and wisdom, among many other things. They tend to believe intelligent people act or are a particular way… and confuse liking a person with thinking he’s intelligent.
As a kid, I was exposed to some type of toxin and my cognitive development was delayed. Nevertheless, later in my childhood I took an IQ test and scored about average to low on the verbal section (reflective of lowered crystallized intelligence), while breaking ceilings of many of the performance sections… giving me an overall IQ in the mid 120s. Since I believe my crystallized intelligence has had some time to catch up, I’ve often been curious about what my IQ is now that I’m an adult.
So I went and sat for a Mensa admissions exam. It was a frustrating experience. For sure, I knew how to solve basically all the problems but I spent some time second-guessing and I feel sure I must have made quite a few dumb mistakes (I know myself well). To be honest I was also a little overconfident and took the tests on an empty stomach, without much sleep, and in a general haze. But what annoyed me the most by far was the strict timing. I make a ton of mistakes when I’m in a rush because I tend not to think carefully about what I’m doing. I’m sure if you gave me a test of 10 arithmetic problems (along the lines of 5 + 2) to do in 20 seconds, and I believed it was a very important test, I would miss quite a few of them. But this does not mean I don’t understand basic arithmetic. I can do advanced math (group theory, calculus of variations), but I can’t do strictly timed arithmetic problems… and that shouldn’t mean I’m stupid.
On the contrary, when I found out I didn’t get a qualifying score for Mensa, I went and took a practice GMAT just for the heck of it (a score in the 95th %-ile qualifies you for Mensa). I more than exceeded Mensa 95th %-ile cut-off (got a 99th %-ile… there’s a big difference between these percentiles in terms of IQ, a 95th %-ile in the general population means an IQ of around 125 and 99th %-ile an IQ of 137) with some time left. This is without any practice. And it’s on the official GMAT practice test so it’s probably pretty close to what I’d get if I took it for real. So, here’s the question: do you think IQ tests place too much emphasis on time?
If you read theory on IQ, you will come across the idea that it’s basically speed of processing and working memory. And I agree with this general pattern, but I think they are missing a crucial element which is something along the lines of “error rate.” Think about this practically, you might have a person who has a 5% error rate. He is a fast processor and has a fairly large working memory. However, when a task starts to become complex, his errors are compounded, so he’ll have to redo the complex task several times before he (by chance) gets the right answer. So it might seem like he’s “slow” but in fact he’s a fast processor. Nevertheless, the difference between this guy (guy A) and another guy (guy B) who has a smaller working memory and is a slower processor (but who nevertheless tests at the same IQ level) is that guy A will get more complicated ideas than guy B, because he will be able to process more complicated ideas than guy A… he’ll just have to be more careful and “rethink” things more often.
I’d like your input on this “three factor” model
Oh and another thing. People often cite, in defense of this idea, the fact that people that score higher on untimed IQ tests tend to get done with them more quickly. Sure, this is true, but on an untimed IQ test no one feels rushed and so people will take all the time they need and then check over their answers (and people with high “error rates” will catch their mistakes here). In this kind of situation more intelligent people will naturally “get” the easy problems faster, but if you give them only easy problems and force them to go fast, their error rates will likely increase and people who already have high error rates will start selecting all the wrong answers. Many intelligent people I’ve talked to complain about missing all the easy questions and getting the hard ones right. It seems like that type of thing is an artifact of our collective notion of what intelligence is.
Oh and another thing. People often cite, in defense of this idea, the fact that people that score higher on untimed IQ tests tend to get done with them more quickly. Sure, this is true, but on an untimed IQ test no one feels rushed and so people will take all the time they need and then check over their answers (and people with high “error rates” will catch their mistakes here). In this kind of situation more intelligent people will naturally “get” the easy problems faster, but if you give them only easy problems and force them to go fast, their error rates will likely increase and people who already have high error rates will start selecting all the wrong answers. Many intelligent people I’ve talked to complain about missing all the easy questions and getting the hard ones right. It seems like that type of thing is an artifact of our collective notion of what intelligence is.
Oh yeah… furthermore, just to clarify even more: Someone who’s a fast processor might do (relatively) simple problems faster with say, a modest error rate of 10%. If you look at the way IQ tests are normed, that already lowers a person’s IQ score. But the punchline is that fast processing people with high error-rates would learn to make up for it by being more careful and doing tests more slowly. You can’t know for a fact that they’re not double checking their answers… it’s likely that they will do so out of habit… even if it lowers their IQ score. If you check the speed of simple calculations you will probably also find that they do them faster (and more accurately) because they double check their work and work quickly (which explains the correlation between the time taken to do simple problems)… but as problems become more complicated and involve several steps, it will take a longer time to double check and find errors. There are studies confirming timed tests measure g less.
Comments are closed.