Most IQ tests only go up to 145. Anything very close to 145 may mean you "ceilinged the test." In other words, your IQ could be even higher, but the test can't measure it beyond that.
Very high IQs can only be accurately assessed at an early age by a qualified professional. So many people will never have an accurate score.
> Most IQ tests only go up to 145. Anything very close to 145 may mean you "ceilinged the test." In other words, your IQ could be even higher, but the test can't measure it beyond that.
This has always bothered me about IQ tests. I mean, how smart is the person who comes up with the test? If your IQ is 120, how could you come up with a question that verifies someone's IQ is 180? Maybe the 180 IQ person would have an answer that is different than yours for reasons that you didn't think of.
So, when I was involved in this stuff, this was my best understanding:
1. Tests are tools. By themselves, they tell you nothing of real value. They are useful in the hands of a qualified professional.
2. Tests are game-able. They try to keep questions and tests under wraps because studying for the test or memorizing answers invalidates the test results.
3. Bright kids are horrendously hard to assess, especially if they are 2xE. They have trouble giving the canned "correct" answer.
This is something my son calls "guess the teacher's password." In other words "what inane thing do I need to parrot word for word to get the teacher off my back."
Example from email list discussion:
"Who discovered America?"
Teacher's password: Columbus.
Gifted kid's reply:
"I'm not sure. Columbus was trying to find India and thought he had, which is why Native Americans were called Indians. But it wasn't called America until years later when Amerigo Vespucci traveled to the continents now named after him...(blathers on about more history)...um, what was the question?"
4. Only certain tests can be used to assess IQs above 145. They were having trouble developing and validating an updated test for the higher IQs.
5. You basically can't be accurately assessed by someone with a lower IQ. Smart kids are routinely given hell by adults of more average intelligence who interpret their behavior as bad behavior and can't accommodate them.
> Bright kids are horrendously hard to assess, especially if they are 2xE. They have trouble giving the canned "correct" answer.
Canned correct answers are only required for knowledge questions (like your example), which IQ tests avoid since knowledge and intelligence are different things.
> Only certain tests can be used to assess IQs above 145.
Most modern IQ tests report up to 160; accuracy (as assessed by test-retest consistency) at 145+ isn't great, though, on any.
> You basically can't be accurately assessed by someone with a lower IQ.
That's baloney.
> Smart kids are routinely given hell by adults of more average intelligence who interpret their behavior as bad behavior and can't accommodate them.
Smart kids routinely are bored and engage in actual bad behavior in situations not optimized for them, but that doesn't prevent adults of adequate but non-exceptional intelligence from being able to apply intelligence tests to them and get the same results as any more intelligent adult would. IQ tests are not limited by the IQ of the test adminstrator.
That answer to the America question is just a student being obnoxious and attempting to show off. They should have enough awareness to know the correct response.
In most cases why would you bother trying to accurately gauge a student over 145? A student with an IQ of 145+ has very different educational needs compared to a 100 IQ student. Do you need to know if it's 145 verse 155?
They should have enough awareness to know the correct response.
This was an actual anecdote from a discussion with a parent of the child in question. The child in question was probably between four and seven years old. They were being assessed by a professional one on one, not in a classroom setting. They were neither showing off nor in possession of the social skills you assume they surely had to have.
I will add that imposter syndrome is common among adults because it's quite hard to figure out where we stand in relation to other people, so smart adults frequently feel they must be "average" because this is "normal" for them. Children are at substantial disadvantage compared to adults.
Do you need to know if it's 145 verse 155?
Possibly. The standard deviation for IQ tests is just 15 points.
More importantly, a test that only goes to 145 cannot distinguish 145 from 180. Both will likely score 145 or close to it.
The use of the word teacher had me confused, as I'd normally assume a classroom setting for teaching student relationships.
I know what these kids go through. I was that kid myself. I know I was very aware that I wasn't "normal". You're correct that in the age range 4-7 a child wouldn't know to give the "expected" answer. It won't be much longer though until that child can use his/her intelligence to work out the expected response. At my school, it wasn't socially acceptable to show intelligence so maybe I learnt to hide it earlier than most.
I don't know my IQ. I've had to take tests a few times as an adult for work and every time I land "most likely somewhere in the top 0.1%" according to the psychologists. I don't really feel it matters if my tests show 145 or 160. A large school with 800 kids is only going to have a single child in that range.
You can either move that child up some years, or you can give them a customised workload. A child that is 180 IQ is very different from a child that's 145 but the approach of attempting to find them an appropiate workload would be a good start.
"Guess the teacher's password" is an expression my son uses. It's based on his experiences attending public school for a few years. (I pulled him out to homeschool him when he was eleven.)
Using it wasn't intended to suggest a classroom setting for the story in question, just a standard answer that someone more knowledgeable about history might not spit out as The Answer without being inculcated to do so.
There are more than 10x as many people at 145 than 155, that ratio is larger than between 135 and 145, because the curve is exponential. If we’re talking about how to “best” educate someone who’s abnormally smart, and if we agree with you that someone at 145 needs to be taught differently than someone at 100, it certainly might matter to know if it’s 145 or 155.
1. Time. Each test question have 1 minute to answer. The test author took 4 hours to write the question (before having an organization put it through a battery of tests and analyses to validate it).
2. Group effort. Test authors work in big teams; many brains makes for easier insight.
3. Children's IQ is normalized to their age scale; a 7-year-old with IQ 140 isn't cognitively equivalent to a 35-year-ld with such IQ.
IQ tests are timed. IQ test questions aren’t something that normal people can’t comprehend, they’re just hard to do quickly. It’s easy to ask hard questions without being that smart, just like it’s easy to time the winner of a sprint even if the winner can run faster than you.
This isn't how IQ tests work. IQ test results are measured relative to a population. In particular 100 is the mean with a standard deviation of 15. So if you score 115 on an IQ test it doesn't mean anything in a vacuum. It means that relative to the population who took the test you scored better than ~84% of people. An IQ of 130 means you scored better than ~98% of people, and so on.
This is what led to the Flynn Effect [1]. It was the initially extremely surprising observation that raw IQ scores over time were increasing. In other words somebody who scored a 100 in 1970 was actually scoring substantially higher than somebody who scored a 100 just a few decades prior. Not so fun fact: starting sometime around the mid nineties, raw IQ scores started declining, at a substantial rate, in many and potentially all developed nations. In other words somebody who scored a 100 in 1995 is scoring worse than somebody who scored 100 in 1975. The reason for this is unknown and controversial (as all things even potentially genetic nowadays), but the effect persists even when controlling for things such as immigration.
Because IQ tests are standardised rigorous tests. 145 are three effin standard deviations up, where 130 means 99th percentile. 145 is 99.9. There simply are not enough people up there to make reliable tests. WAIS (the most used one) breaks down around 135.
They have different tests with better accuracy in different ranges. I didn't know this until relatively recently and never understood why I was given 3 tests in as many months. If they think you're slow instead of bored to death they may start with a test looking for developmental delays.
Per Wikipedia, most IQ tests have a highest-reported standard score of 160 [0]; that's also consistent with the highest range for Stanford-Binet 5th Edition and Kaufman Assessment Battery for Children topping out at 160 (the SB5 top range starts at 145.)
Even if that's the highest achievable, IQ tests designed to be usable for everyone tend to start getting really unreliable above 135 or so. Above that level, you're getting into the territory where test takers basically answer almost all of the reasonable questions right and where those few wrong answers necessarily start having much larger effects on the final score. Because of that, other issues start dominating the results. Random variance, sloppiness, mechanical issues in timed tests, how well you slept, whether there's slightly bad/ambiguous questions, whether you've seen that type of question before, whether you're good under pressure, cultural differences, etc.
For example, one serious official test I did included a timed test where you had to decode a string of symbols to text based on some key. I ended up being fast enough at decoding that I was effectively limited by how fast I could scribble down the decoded texts and turn the page back to the code, not by the mental calculations. I imagine others would have had similar issues, making this particular test at best capable of indicating a wide possible range of mental ability, a range which should be particularly inaccurate around the higher ranges as the ratio mentalTime/mechanicalTime decreases.
There also just aren't enough high IQ people to really effectively study the accuracy of these tests above those levels. There are no good baselines to compare these people and no real ways of checking whether the differences actually matter in any meaningful way.
> IQ tests designed to be usable for everyone tend to start getting really unreliable above 135 or so.
Afaik, that's not a “designed to be usable by everyone” thing; there's test-retest consistency gets progressively lower beyond that point on IQ tests, period.
> There also just aren't enough high IQ people to really effectively study the accuracy of these tests above those levels.
If that was true, we wouldn't have reliable information that accuracy breaks down in that area, which we do.
We can study the accuracy in that range, we just haven't found a way to fix it.
How is it possible for anyone to have a valid very high IQ, since IQ is a percentile test and there's no way to know how many very high IQ people were never properly tested?
Very high IQ is fundamentally flawed because IQ is resistant to precise measurement.
Social systems, like kindergarten, can tell if someone is smart or average or very not smart. Society generally only tests those at the outer boundaries to verify and classify them into groups.
You're right, someone who seems average (or totally unknown to social systems), but is actually super genius, might not get tested. It's the "genius in the rice paddy" argument and yes, those exist. Srinivasa Ramanujan was one.
That doesn't mean IQ is fundamentally flawed. You don't need to know if someone is 99th percentile or 99.0004 percentile to know they are probably the smartest in the average room of a hundred people and that's what we use IQ to determine.
Sorry to be nitpicky but this stuff is interesting to me. I would think someone in the 99th percentile would have a 99% chance of being smarter than another random person. In a room full of 100, to be smarter than all 99 other people would be .99^99 or only around a 37% chance of being the smartest in the room. I may be off on this, as I am definitely not in the 99th percentile.
Very high IQs can only be accurately assessed at an early age by a qualified professional. So many people will never have an accurate score.