IQ test

[Kali]

I downloaded an IQ test application in my cellphone just for fun. My result was 133 (deviation 15). I researched a bit and it turned out that the test used in this application is one of the official tests designed by Mensa (I found the same test online), but I didn't know about the 40min time limit for finishing the 39 exercises, as the android application don't have this time restriction and said nothing about it. I don't know exactly how long it took to me to finish it, but I think it was less than an hour (I spent most of the time in the last 4). I want to know if this time restriction is very important for the scoring, as I'm kind of surprised with my result. Also I can't make the same test again because I'll remember some of the exercises so it will take me less time to do it. Anyone knows another good test so I can get a proper and valid result for re-checking the one I obtained? The one I did is this: http://www.iqtest.dk/main.swf

I didn't make this post for presuming, also I'm kind of skeptical about this... As far as I know 133 is really high, and too much for me considering what I think my real intelligence level is, so the time restriction must be very important for the test even when I think it didn't took me much more than 40min. But I'm considering also that without knowing about a time limit, I made it with no hurry and very relaxed, so maybe I can get a similar score anyway with another tests.

But if I'm really 133, then I think some people in this forums must be around 200 

If someone here took some of this tests and wants to share their results, I'm curious... 

[hobold]

An IQ test is only a "test" and not a "measurement". The resulting "point score" is a unitless number, i.e. it does not track any quantity, but is a statistical comparison to a(n idealized) "representative" sample of humanity.

That means:

- the time limit is important, because it was in effect when the reference scores were sampled

- the score of one person can vary fairly widely across different IQ tests, because each test covers only part of the breadth of human mental ability

- repetition of the same test does indeed skew the statistics so far as to make the results useless


I cannot really answer you actual question if you should believe your scores. But I can tell you that is it fairly common for intelligent people to underestimate themselves. This has to do with the fact that intelligence doesn't just mean that you are aware of and can process more information. It also means that you are more aware of the limits of your knowledge and skills. You might want to google for "Dunning-Kruger effect" for a fascinating journey to the psychological limits of human knowledge.


Additionally, the ability to be critical of oneself (i.e. to recognize the flaws in one's own work) occurs more often in gifted people. Or rather the other way round: when you are good at realizing your mistakes, you have much more of a chance to learn from these errors. And so an intelligent and self critical person will tend to learn faster and learn more and discover more new knowledge, and might ultimately produce more interesting results. It's not that good for your happiness, though, to always see your failures more clearly than your successes ...

[eiffie]

It may be fairly accurate Kali I got 112 and that is about what I remember testing at years ago. Your mind is wired for these pattern matching and rotation skills.

[fractower]

Hobold: I can tell you that is it fairly common for intelligent people to underestimate themselves.

The Dunning–Kruger effect. Though it is usually expressed as the corollary due to the greater negative impact and potential humor value.

http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect

[taurus]

I want to know if this time restriction is very important for the scoring

Afaik it is important for scoring. For ordinary humans, it isn't possible to solve everything in time - that's part of the test.
Additionally a serious IQ test should consist of different sections (practical, linguistic, math etc).
I woudn't take this too serious. A wise man oce said, the IQ measures the ability, to solve an IQ test... 

[M Benesi]

  Just remember, subtract 20 points from your score if you buy the detailed analysis of the testing results.

[lycium]

The Dunning–Kruger effect. Though it is usually expressed as the corollary due to the greater negative impact and potential humor value.

Sorry to nitpick logic terminology here but that's not what a corollary is

@Kali, who here should have the 200 (SD 15) IQ? I didn't see William Sidis posting in here...

[Kali]

Well, many thanks for the info and comments. I also refuse the idea of something such complex as our brains and intelligence to be measured by 39 pattern-recognition exercises. Maybe is just a way for evaluating raw potential capabilities, trying to leave out of the calculation the previous knowledge or specialization on any area of the test subject.
I've read something about multiple types of intelligence, and maybe I'm wired for solving this kind of pattern matching as eiffie said, but I can say that I'm not really that good using my alleged intelligence in many aspects of my life... but ok, maybe this could be something related more to my psyche than my IQ  .

@lycium: ok, but you can't deny there are some pretty brilliant minds in here. I don't want to mention anyone because the ones not mentioned could feel bad  .
But I can say that if eiffie got only 112, something must be wrong with this IQ tests

[kram1032]

An IQ test indeed highly depends on the sample of people it's normed on. As such, the higher (or more generally, the more distant from IQ100) the result is, the less accurate it is.
Furthermore, IQ tests are very weak proxys of intelligence, focusing way too much on analytical procedures.
Also, if you're doing a test in a language that isn't native to you, you'll automatically be scored worse.

Mostly (not always), IQ tests test how well you generally do at a typical school, which, if you think about it, hardly represents true intelligence. Lots of people you might call geniuses did rather bad at school.

The more recent you get with research, the more often you'll find results that suggest, that IQ tests are pretty much meaningless and don't test for the core factors behind those tests.

To answer your question, though: The timing of an IQ test is extremely important. Depending on the test and time, just a single sneeze might cost you 4 IQ points for being left with less time.

I have done various IQ tests with varying results before. IIRC I always was in the range between 120 and 165. (165 was the highest I've got on one).
The sheer variation between those tests suggests that they are uncomparable. If you really want to compare your IQ to someone else's, you'll have to do so under precisely the same conditions, with the same test, the same timing and with both being equally awake. The latest one is pretty much impossible to get right but can have severe effects on the score.

[hobold]

A wise man oce said, the IQ measures the ability, to solve an IQ test... 

This is the only claim about IQ tests that can be made with certainty.


Another fun fact: the scale for IQ test results is not consistently normalized. Sure, 100 is always the average. But some tests value one standard deviation out as 110 points, while others value the same "raw" result as 115 points. So the same number 130 can mean either two or three standard deviations from the mean. And the difference between those two points on a gaussian probability distribution is ... huge. The air quickly gets thin up there.

Extreme abilities in an isolated small skill set are not all that uncommon. So as I wrote earlier, depending on the type of test, scores can vary a lot. If the test only exercises the few genius parts of your brain, you'll score very high, even if all your other mental skills are closer to average.

And lastly, IQ tests do not convey any information about the style, the quality of one's thinking. The most important part in my not so humble opinion is intuition and creativity. A well educated quick thinker may be able to, say, solve a large number of mathematical computations with brute force at breathtaking speed. But a more ingenious mind might recognize a geometrical pattern in those formulas, and intuitively grasp that the solution has some meaning as a distinctive shape. An IQ test cannot distinguish between those two hypothetical people. (And let's not even get into issues of motivation, i.e. why we do what we do, regardless of whether we are "naturally" good at it.)

[kram1032]

Some of the tests I tried so far, I did together with others, back in school.
Usually people say, you top out intelligence at ~26. There is a lot of debate on that value though.
Either way, as said, back then I was in school. And not a single one of my classmates scored below 110 on either of those tests. The scores varied between 110 and 170.
Now have a random selection of 20 people and take their IQs. How likely is it that not a single one of them turns out to be below average? How likely is it, that every single one is above average? Both clearly indicate that the "100", those tests were normalized to, are not *actually* the population-wide average.

A quick probability proxy would be having a coin land on heads 20 times in a row, about one in a million.
And I can assure you that

• some of my class mates weren't the brightest
• our class was not a class in a million.

Furthermore, imagining that, what was the age back then, 13-14?-year-olds already have IQs up to 170... What would those end up getting when "on their high point"? 200+?

Clearly, if you want a "true" IQ test that actually might have something like a vague distant meaning, you'll need to go to some sort of specialist / psychologist. And even then, IQ tests are rather useless.

To some extend, I'd even say they're dangerous, since they assign a number to you that doesn't have a true meaning but is perceived to be marking you as genious or retarded, giving rise to a lot of prejudice and wrong expectations.

[David Makin]

Some of the tests I tried so far, I did together with others, back in school.
Usually people say, you top out intelligence at ~26. There is a lot of debate on that value though.
Either way, as said, back then I was in school. And not a single one of my classmates scored below 110 on either of those tests. The scores varied between 110 and 170.
Now have a random selection of 20 people and take their IQs. How likely is it that not a single one of them turns out to be below average? How likely is it, that every single one is above average? Both clearly indicate that the "100", those tests were normalized to, are not *actually* the population-wide average.

A quick probability proxy would be having a coin land on heads 20 times in a row, about one in a million.
And I can assure you that

• some of my class mates weren't the brightest
• our class was not a class in a million.

Furthermore, imagining that, what was the age back then, 13-14?-year-olds already have IQs up to 170... What would those end up getting when "on their high point"? 200+?

Clearly, if you want a "true" IQ test that actually might have something like a vague distant meaning, you'll need to go to some sort of specialist / psychologist. And even then, IQ tests are rather useless.

To some extend, I'd even say they're dangerous, since they assign a number to you that doesn't have a true meaning but is perceived to be marking you as genious or retarded, giving rise to a lot of prejudice and wrong expectations.

Please describe which scale you are using - the US appears to use a system where 170 represents "genius" level, but the system in the UK uses 130 (or maybe 135?) - of course I think they are equivalent but with different difficulty/scoring.

I'm apparently UK 132-136 for anyone interested - based on the public tests done a couple of years ago by the Beeb (which covered all the "usual" IQ question types) - the largest result made public of anyone who took part was 146 I think.
My brother (and his wife) are Mensa members - that's how they met I think !!
Also I have mild Asperger's (now considered low level Autism) which is apparently relatively common with higher IQs - though (IMO) I'm nowhere near as bad as Sheldon but worse than Leonard
Unfortunately in my case I'm severely lacking in formal academic education since I've never completed anything beyond UK "A" levels and I've been trying to make time to get up to speed on Degree-level fractal related math but to be honest something always gets in the way.
A couple of years ago I asked my brother to get me a degree-level math text book relating to fractals and he got me "Chaos, Dynamics and Fractals: an algorithmic approach to deterministic chaos" by J.L.McCauley (Cambridge press) - it might as well be Greek, I don't even know the mathematical syntax used in the first few pages for certain.

Unfortunately my brother can't help - he's in Mensa but he didn't do Maths even to "A" level.

[Tglad]

IQ tests are irrelevant, they're more or less completely meaningless. 
Ability to quickly solve little pattern recognition tasks is nothing to do with intelligence (whatever that is anyway), which is why no academic institutions use them, nor do any research industries give them any credit. Einstein most probably would have had a lower score than Sylvester Stallone.
The only group that thinks IQ tests are valid are MENSA, but they would since it is self-serving.

[JohnVV]

who knows

All i know was 35 years or so back when i was young( pre-teen) they said i was 165
now that was before the drugs and booze -- man i liked that needle
and 35 or so years back so who knows now
and the tests HAVE changed

PS. the mensa challenge was very easy ,but that was when i was 17

[Alef]

It's not only about multiple intelligences. Probably IQ is used, becouse in certain jobs it helps to get rid of dumb .

Well IQ tests alsou are culturaly biased. Englishman after reading the task intuatively  would understand what is asked. Russian at first would need to think, what is asked to do. And only then do the task. So he will have lower IQ score.
I have few indian friend with good english, but sometimes I need to think a lot, what he realy meant. Same words in different cultures can have different meanings. In darkest racist corners of internets (like wikipedia) informs that chinese are exeptionaly good at IQ tests, but this could reflect that they are more disciplined.

It alsou gives number of your worth.