Log in

No account? Create an account
entries friends calendar profile Previous Previous Next Next
IQ - Rachael
I had my IQ tested as a teenager, and it was an important part of my identity for a few years. But when I came to Cambridge and started to meet people who were as clever as me and people who were definitely cleverer than me, I was surprised that they'd never had theirs tested. And in the circles I move in now, it's a topic that never comes up. So I suppose if you grew up in that kind of milieu, with clever friends and family all your life, it probably wouldn't seem a big deal or occur to you to have it tested.

(This is leaving aside the sociological criticisms of IQ. I don't especially want to have that debate, unless you really want to. I'm on the pro-IQ side of the debate, FWIW. But this post is mainly about the practice of measuring it, and its distribution, rather than its meaning or validity.)

I was just thinking about it because my mum got in touch with me to say she was watching a TV program where they seemed to be claiming a child had the highest ever IQ, and she said she thought mine was higher than that, and I said that yes it was (and dug the certificate out of the loft just to be sure), and that I'd known other people with IQs higher than that too, and that I'd read about people on the internet with even higher ones. I think mine is probably only average for the social group I'm now a part of. But if it's not customary among very intelligent people to get it tested, the statistics at the top end (which is necessarily a narrow tail) might be distorted due to measurement bias, and the famous very-high-IQ people (like Marilyn vos Savant, etc) might not be as much outliers as is currently thought.

I'm wondering whether my impression that "it's not customary among very intelligent people to get it tested" is accurate. You lot are part of the social group in which I consider myself intellectually average; did you ever have yours tested?

Tags: , ,

25 comments or Leave a comment
toothycat From: toothycat Date: July 13th, 2015 02:21 pm (UTC) (Link)
I (S) had mine "tested" when I was 12 or so.

I encountered a Mensa ad in some publication or other - more serious than a comic, but pretty sure it wasn't a newspaper; surrounded by very much the "innovations catalogue" style of advert pages rather than the "X-ray specs" style or the "buy a PC for your small business" style. I was already familiar with the name Mensa from the puzzle books you see around, so applied to take the tests.

After a few interactions I came to the distinct impression that whatever one might think of the tests, the organisations typically plying them aren't actually useful or interesting - they're basically "you pay us money, we make you feel good about yourself" quasiscams, a bit like the "pay us money to be in this "directory of notable people" (i.e. people who paid us money) crowd you regularly see around; the only people that care about the test results are the other Rimmer types prepared to pay to get "BSc, SSc" after their names, no-one else cares and there is literally no purpose to which one can apply either the test result or the memberships other than meeting this crowd of other people who have paid to do so.

So I felt vaguely embarrassed and duped for a while (though I can now neither remember nor work out what I *expected* to happen, exactly!) then proceeded to forget the whole experience as rapidly as possible.

Since then, I've never directly seen any body other than Mensa ever mention the possibility of widespread IQ tests for any reason. Every so often I see references to research results, so I assume some postgrads are doing studies somewhere, but I've not come into contact with any of those, and never seen any more organised efforts to test the population.

I gather your experience is different?

Edited at 2015-07-13 02:24 pm (UTC)
woodpijn From: woodpijn Date: July 13th, 2015 03:33 pm (UTC) (Link)
I also took the Mensa test, I can't remember where I heard about it - it might have been a similar magazine ad. I was interested in the result (I think I put it on my CV for a couple of years, like, during sixth form, which I now feel embarrassed about for the Rimmer-ish reasons you mention); and I was also interested in the societies within Mensa, for the purpose of meeting interesting clever people to discuss interesting things with (not necessarily face-to-face - the societies operated by magazine distribution (which were typewriter-typed and photocopied, in the late 90s)), and possibly for finding a boyfriend :) Two things happened soon after that, namely access to the Internet and getting into Cambridge, which were better for both purposes. (I assume Mensa now runs its societies as online forums, but there are more and better and self-selecting online communities for discussing interesting things.) So my family only paid the membership fees for a year or so and then we stopped bothering.

But if it wasn't for the internet, and if I couldn't go to university for whatever reason, I might still think it worthwhile to be a member.
geekette8 From: geekette8 Date: July 13th, 2015 03:52 pm (UTC) (Link)
Very similar experience to yours, except that it was my parents who pushed it rather than my own interest. They had this idea that "Mensa would look good on your CV" which at the time I believed but now of course hahahahaNO.

I did the test when I was 17 or so, can't remember what the numerical result was except that it was comfortably enough to get into Mensa, and belonged to Mensa for a couple of years until I went to university and realised how pointless it all was.

I went to a couple of meetups in Shrewsbury (where I grew up) and in Cambridge (where I went to university) and didn't feel I had anything remotely in common with any of the other people at either of them. My most significant recollection of it all was that the woman who hosted the Cambridge meetup I went to made and served a FISH CRUMBLE which was just a horrific concept to me and I wasn't sure how to politely decline given that I had already accepted that I would like some lunch please before I knew what the lunch actually was.

I did also join a couple of the special interest groups (can't even remember which ones now, other than that one was the computing one) and enjoyed the typed-and-photocopied magazines. I even wrote an article about girls losing interest in computer science during their school years for the computer SIG magazine, although I suspect I'd find it horrifically embarrassing to re-read now so I can only hope that no-one has ever bothered to put them online.
ceb From: ceb Date: July 13th, 2015 08:43 pm (UTC) (Link)
I remember those ads, they were all over the place when I was a young teen, and I was kind of obsessed with logic problems and the like at that age.

ISTR I sent off for the test but never got it marked (probably partly because it cost money and partly because it seemed kinda frivolous). Or maybe I never sent off but we had some kind of puzzle-book version instead? No actual number, anyway.
emperor From: emperor Date: July 13th, 2015 02:25 pm (UTC) (Link)
I have no recollection of ever having my IQ tested.
gerald_duck From: gerald_duck Date: July 13th, 2015 02:54 pm (UTC) (Link)
I had mine tested when I was 8 and people were worried at my being bored of school. Then the school I was at when I was 11 tested me and found me off the top of the 45-135 test. At 13, there was an IQ test as one of the entry exams to another school — I came top overall amongst the applicants, but never found out my IQ score individually.

I've never taken an IQ test for my own edification, though, and that test when I was 11 is the only result I've ever known.

Mensa explicitly tried to recruit me after I was chosen for the UK's IMO team, but I declined.

My experience has been that people with high IQs tend to be good at things, and can build their identity around the things they're good at. People only seem to fixate about IQ if they lack aptitude for anything. Or, more often, pushy parents fixate on IQ. (-8
woodpijn From: woodpijn Date: July 13th, 2015 03:33 pm (UTC) (Link)
That makes a lot of sense.
From: khoth Date: July 13th, 2015 03:44 pm (UTC) (Link)
I had mine tested, but only because a family friend happened to be a child psychologist. I don't remember how old I was, or the result (except that it was high).
From: ptc24 Date: July 13th, 2015 05:51 pm (UTC) (Link)
Mine, when I was about 9 or so, my parents had taken me to an educational psychologist, and I got it tested then as part of the evaluation. I have the report and everything, with a bunch of subtest scores (very interesting - apparently I have a "spiky profile", there's a bigger-than-usual difference between my strong and weak areas). Wechsler Intelligence Scale for Children - Revised.

However, the IQ was only a small part of the evaluation - there were various difficulties which in retrospect associated with Asperger's, although the closest the report got was "Peter does have rather more [autistic traits] than most children but the label of autistic would be an inaccurate appelation" - and that was the main reason for going and seeing him.

I also recently had a short-form IQ test, which I didn't get the results from, as a part of someone's research.

It would be interesting to know where you'd put the average IQ in our circles. Top 2% is 130, top 0.1% is 145, I think I'd put the average somewhere in between. Not least because mine is/was[1] in that range and if anything I see myself as a bit above-average for our crowd.

[1] Childhood IQ is a bit odd as it is standardised for age.
woodpijn From: woodpijn Date: July 14th, 2015 07:35 am (UTC) (Link)
I'd put it quite a bit higher than that because mine was quite a bit higher than that and I see myself as average in our group.

But I'm wondering if my result was a bit distorted by the childhood adjustment you mention and/or the speed issue livredor mentions.

Edited at 2015-07-14 07:36 am (UTC)
From: ptc24 Date: July 14th, 2015 08:10 am (UTC) (Link)
And of course there's good old Feynman, weighing in at about 125 or so. Although apparently the test he took was biased towards verbal ability; Feynman could write well and amusingly, but that wasn't what he was noted for.

Apparently on one of my subtests I was held back by my slow speed with a pen; on another subtest I maxed out, but whatever. On the level of percentiles (easier to interpret than an IQ score), the result put me in the top 0.4%, which sounds about right. On the other other hand, specialisation has let me trade on my strengths, which given my "spiky profile" might mean my IQ could be lower than my apparent intelligence in these circles suggests.

Other distortions: there's the issue of test-retest reliability - apparently, the correlation between two IQ test scores taken by the same person on different days is about 0.8, although I can't remember whether this is r or R2. Incidentally, this is the same correlation between the scores of identical twins.
woodpijn From: woodpijn Date: July 14th, 2015 06:29 pm (UTC) (Link)
Another bizarre complication is that the Mensa test certificates say that your score is X and that puts you in the top Y % of the population, but they don't seem to support non-integers for Y. So even if X is a lot higher than 99th percentile, it says "top 1%". This is the case on my certificate and the ones that come up on a Google image search.
toothycat From: toothycat Date: July 13th, 2015 08:45 pm (UTC) (Link)
I don't think I had mine tested, but I might have done. If I did, it would have been as a child, it can't have made much of a impression on me and I've forgotten any details ^^;
pigwotflies From: pigwotflies Date: July 13th, 2015 09:16 pm (UTC) (Link)
Nope. I would be curious to find out, but not enough to actually investigate how and spend money on it.
woodpijn From: woodpijn Date: July 14th, 2015 07:36 am (UTC) (Link)
Maybe you could volunteer for some study which includes an IQ test as part of it. Then they might even pay you.
andrewducker From: andrewducker Date: July 14th, 2015 08:21 am (UTC) (Link)
That was how I got mine done - I helped out as part of a study where I had an MRI taken, and an IQ test done.

I was off the top of the scale, but it only went up to 135, IIRC. (I was 34 at the time)

"Getting your IQ tested" certainly wasn't a thing when I was a kid. I was aware of IQ and Mensa, but I didn't know anyone who was actually doing it.
angoel From: angoel Date: July 13th, 2015 09:22 pm (UTC) (Link)
I've had mine tested at a couple of points in the past, generally reaching a figure of just below 140.
livredor From: livredor Date: July 13th, 2015 10:19 pm (UTC) (Link)
I think I probably had my IQ tested and scored highly when I was 5 or 6, because I ended up in a LEA gifted kids' programme for a while. But the adults involved intentionally didn't tell me the number. Since then I've never formally done a certified IQ test, but whenever I take that kind of abstract reasoning test, aside from obviously meaningless internet quiz ones, I come out somewhere over the 99th centile. (Eg I did a fairly serious career counselling psychometric test in sixth form and I was off the top of its measuring scale for every skill except spatial reasoning where I just fell below the 99th centile.)

Honestly I don't think even formal IQ tests are very good at distinguishing 99 from 99.9. The bias at the top end may be partly because people who are highly intelligent usually have some more meaningful way to demonstrate it (eg I have a science PhD, which most people assume means I'm pretty bright), so as you say people who would score highly might well not get tested much. But I think there's also bias because anything above 130 ish is really hard to measure sensibly. Among people who get basically all the answers correct I think the tests end up measuring speed, which depends on a lot of factors.
gerald_duck From: gerald_duck Date: July 13th, 2015 11:47 pm (UTC) (Link)
A random question that occurs: how easy is it to design an IQ test that tests for an IQ higher than one's own?

It may be that those capable of creating an IQ test that discriminates accurately between people above 130-ish have better things to do…
woodpijn From: woodpijn Date: July 14th, 2015 07:46 am (UTC) (Link)
Yes, very good point (obvious in hindsight but hadn't really occurred to me). Reminds me of the problem in fiction writing of trying to write a character more intelligent than oneself.

I'd imagine there are probably some superintelligent people working in IQ research, just because people's interests are very varied; BICBW.
From: ptc24 Date: July 14th, 2015 10:56 am (UTC) (Link)
Possibly not all that hard. I was about to suggest "get a panel of people, each top-rate in their own subtest" but it's probably a lot easier than that.

Some subtasks are obviously easy to make arbitrarily hard. Vocabulary, for instance, you don't need a genius to design a good test, you need a dictionary. Mental arithmetic is another good one. This doesn't obviously scale to more "puzzle-like" things like Raven matrices - those puzzles where you have a 3x3 grid with a cell missing and have to fill in the cell.

That said, I think it's easier to come up with the answer first and then convert it to a question than vice versa. I know from my own experience that when playing Zendo it's easy to come up with rules which turn out to be very hard to guess.

Think about crypto - large parts of crypto are about things where if you start with the answer it's easy to get the question, but not vice versa. Like public-key cryptography.

The main problem - having generated these problems, how to standardise - i.e how to convert raw test results into IQ scores. If your test is meant to distinguish people in the top 0.001% (that's IQ 164 if my understanding and maths are correct), then you need a big test sample - 100000 people - in order to set the norms out that far. (ooops, hang on, there's a comment right below mine saying the same thing...)

Edited at 2015-07-14 10:57 am (UTC)
From: khoth Date: July 14th, 2015 10:58 am (UTC) (Link)
It's probably not terribly hard. For example, with pattern-recognition type questions, it's obvious from playing Zendo that coming up with a question is significantly easier than figuring out the answer.

Hah, and someone else brings up Zendo in the two minutes I spent writing this...

Edited at 2015-07-14 10:59 am (UTC)
woodpijn From: woodpijn Date: July 14th, 2015 07:38 am (UTC) (Link)
That's a really good point. I think I scored highly just because I finished the test quickly. (I always finished school exams very quickly too.) There probably aren't enough examples at the top end to calibrate the tests properly.
ilanin From: ilanin Date: July 14th, 2015 04:40 pm (UTC) (Link)
I had my IQ tested at school for a mix of pushy parent/bored problem child reasons; I came out with a ridiculous verbal reasoning score of 176 (which I think must have been an artefact of incompetent age-normalisation because while I am good with words, I'm not that good) and a more reasonable spatial awareness of 148 (not much more reasonable - my spatial awareness is not much better than average in my opinion, so that should be 110ish). (My experiences with MENSA are identical to yours, except I never put it on my CV because...well probably because it would never have come up. I never actually wrote a CV until I was 16 or so and stopped being in MENSA around 14).

mair_aw From: mair_aw Date: July 15th, 2015 11:40 pm (UTC) (Link)

when I was in ?infant school, I was performing badly in school and yet my parents had the impression I was intelligent, so they decided to get me tested to find out The Truth.

Turned out I was quite bright but couldn't be bothered to tell the teachers things I knew they already knew.

I had a friend about that time who was also performing badly but appeared quite bright and I think her parents were inspired/encouraged by mine to get her tested. Turned out she was quite bright but badly dyslexic. At times her written work was completely incomprehensible and teachers could be quite sarcastic. I think having the IQ result reassured her somewhat against those put-downs.

Nobody bothered testing my little sister, as far as I know.
25 comments or Leave a comment