18.01.09

Sorry, you don’t have a 200 IQ

in shorter, writing

You just don’t.  Nobody does.

People like to speculate on the IQ of various celebrities. A popular one recently was Sarah Palin (“likely somewhere between 110 and 115“).  We may also have “learned” that Beyonce’s IQ is 124 (or is it 110?) And nobody can agree on Einstein’s putative IQ, except that it was somewhere between “only” 160 and 250 or more.

Online, I’ve seen the same group of nerds who enjoy self-diagnosing Asperger’s report their IQ scores as 180 or more (200 is a popular number).  There are two possible reasons a person might say this:

  1. They took some fake test on the web.
  2. They are making it up.

Whatever you happen to think about the intrinsic worth and predictive ability of psychometrics, real IQ tests are based on math.  The math expresses how many other people in their sample population achieved the same raw score that you did, by percentile.  Ideally, this sample group is a cross-section of the population that’s representative of the subject’s environment.

The most popular and reliable IQ test for adults is the Wechsler Adult Intelligence Scale, or WAIS, now in its fourth edition.  It was normalized against a sample of 2,200 adults.  Keep that number in mind.

This graph should be familiar to most people.  It’s a Gaussian curve, or normal distribution, of performance on a properly normalized IQ test like the WAIS.  Each color change represents one additional standard deviation from the mean.

picture-30

It’s instructive to look really closely:

picture-31

Each standard deviation in the WAIS is 15 IQ points further from the mean of 100. Because performance is normalized, only 0.27% of those taking the test are expected to fall outside of 3 standard deviations — in the WAIS this translates to scores above 145 and below 55.

If you take the WAIS and achieve a raw score comparable to only the top 0.135% of the original sample of 2,200, it means your performance is measured relative to 3 people.   Score 4 standard deviations above the mean (IQ >= 160) and you’re being compared to just 0.065 other geniuses.  In other words, it’s highly likely that no one in the WAIS sample scored as high as you.  Congratulations, you are “only” as smart as the lowest estimated IQ of Albert Einstein.

Above (or below) a certain threshold, IQ performance is simply noise.  If you extrapolated all the way to IQ 200 (and if you were that smart, you understand why you can’t), you’re scoring a whopping 6 standard deviations above the mean and will have to look elsewhere for your intellectual equals.  Since 99.9999998027% of a normally-distributed group falls within 6 standard deviations, the number of members of your uber-Mensa is 6.  In the entire world.

The fact is, most adults simply do not know their IQ.  Bright children are rarely tested as a matter of course (although some private schools do it).  In general, a child is given an IQ test when their school record is lacking, usually because there’s a disparity in expected versus actual performance.  IQ testing is a good way to reveal that an otherwise smart kid has a particular learning disability.  Psychologists typically do not care whether your IQ is 130 or 145.  They want to know if your non-verbal IQ is high but your reading score is below-average; you may have dyslexia and need special educational strategies to succeed.

But the real benefit of IQ is knowing that when someone quotes you a number and it’s greater than 145, it’s safe to assume they’re not as smart as you.

posted by liza
comments: RSS 2.0
Currently Comments are off, please use a Trackback!

8 Comments

  1. J. Robinson Wheeler said,

    January 19, 2009 @ 4:36 am

    Excellent. Thanks for posting this.

  2. Adam Cadre said,

    January 19, 2009 @ 5:13 am

    Oddly enough, while after nearly 35 years on this planet I still don’t know my frickin’ blood type, I actually did have my IQ tested as a child by a psychologist in Bala Cynwyd. “Stanford Binet,” these records say. Is that any good?

  3. liza said,

    January 19, 2009 @ 2:50 pm

    Sure, the Stanford-Binet is a real test and it would’ve been administered by an actual trained psychologist.

  4. Lucian Smith said,

    January 20, 2009 @ 11:14 am

    I thought I remembered that there were two competing methods of measuring IQ, one of which had a much higher standard deviation, so if you scored 115 on one, the same percentile would rank a 140 or something on the other. I assume the more generous scale was created to make people feel better about themselves.

  5. TERRY FRASER said,

    March 25, 2009 @ 4:37 pm

    The newer tests only go to 135+

  6. Brady Yoon said,

    May 31, 2009 @ 10:32 pm

    Yeah, a 200 IQ is really unreasonable. Pretty much, if someone says that they have an IQ of 145, and they really do, you won’t be doubting them. An IQ of 145 is like a height of 6’7″. Everyone will soon find out that you’re really intelligent. So I guess I’d revise your statement and say that if someone quotes a 145+ IQ and you look at him with a raised eyebrow, you’re probably right.

    I personally have an IQ that is high enough that I hit the ceiling at virtually every standardized test (SAT, GRE for example). To me, the idea that these tests are actually “standardized” is hard to conceive because I’ve always scored at 99th percentile, and I’ve become very used to seeing that 99…I’d say that my IQ is 155.

    From personal experience, this is how you can tell if someone has an IQ of 155: Sense of humor, both depth and breadth of knowledge. You’ll be able to talk to him about anything and he’ll hold an intelligent conversation and hold up his end of the argument. I hate it when people say that intelligent people are bad conversationalists. The difference between smart people and not so smart people in conversation: the smart person will recognize when the other person has said something intelligent, and that he needs to reassess his beliefs or rethink his position. The not-intelligent person will not respond to your statement, but will simply act as if “not listening.”

    A huge complaint that people have is people “who don’t listen.” Listening = understanding. Understanding means intelligence.

    If someone tells you that smart people are bad communicators, that person is not smart, and he’s not a good communicator because he’s spewing BS.

  7. Brady Yoon said,

    May 31, 2009 @ 10:54 pm

    Oh yeah…one more thing. When women say that they look for confidence, sense of humor, that is the outward manifestation of intelligence. Our world is set up so that intelligent people will find life “easier,” and hence life is a confidence builder (don’t believe this? everything in life requires thinking and reasoning at some level, it is because you’re intelligent that you don’t realize how much intelligence permeates every aspect of our world, as those with traumatic brain injury or stroke, Alzheimer’s realize). Sense of humor comes from seeing connections between seemingly unrelated things and connecting them in a new (and funny) way. Why has intelligence developed? Natural selection, yes. But also sexual selection.

    Intelligence was of huge survival value, but it has evolved to such a staggering extent, and that was probably due to sexual selection as well.

    Do you know why women say that they don’t like nerds? It’s because they are TRYING to sound smart. Nerds are people who act according to a stereotypical “smart” person. It’s just like the person who tries to be funny but isn’t. The funny person is like the smart person and the “nerd” is like the person who is trying to be funny. Trust me, intelligent people are well-adjusted, confident, are fun to be around.

    I’m not venting here. I’m trying to say that there are so many stereotypes of intelligent people that apply to the people who try to mimic intelligence rather than actually intelligent people. Smart people don’t try to come across as smart just like funny people don’t try to be funny. Mimicking intelligence is obviously a very valuable evolutionary strategy, simply because it works. Think about what differentiates us from animals, what is our defining trait – intelligence. People keep talking about emotional intelligence, social intelligence, analytical intelligence, etc. Those things are PERSONALITY. And I’m not sure of this, but I’ll make a big bet that “personality” arises as variations in the cerebral cortex, the center of intelligence. Do animals have rich personalities? No.

    Intelligence is like knowing how to talk, or being able to see clearly in full color and depth perception, or being able to walk bipedally. It’s such a defining characteristic of Homo sapiens that it’s easy to dismiss and downplay.

  8. Daniel said,

    November 11, 2009 @ 10:16 am

    I love it when people continually spout misconceptions about what IQ really is. Usually it comes with people talking about “oh well I had a high IQ and high test scores on the SAT” implying a correlation or causation that simply is not there.

    You won’t be able to tell someone’s IQ by talking with them, viewing their interactions, viewing test scores or achievements. The best I’ve been able to understand IQ is it’s more akin to horsepower. IQ means you can get from thought A to thought B quicker, or you might be able to juggle more abstract concepts, but doesn’t imply you ever even made it to thought B.

    Frankly I love IQ because it’s an instant sign when someone touts their IQ as a badge of honor they’re probably not the brightest or at least very arrogant. I like to laugh inside when people quote me their IQ’s.

RSS feed for comments on this post · TrackBack URI