IQ test in high school, 156. SAT 793⁄800 verbal, 783⁄800 math. Cal Tech. Yatta yatta. But that was many years ago. It’s pretty obvious what happened. Timed test. I only finished, in the time, about 2⁄3 − 3⁄4 of the questions, maybe a bit more, I didn’t keep count. (I skipped questions that weren’t popping up right away, thinking I’d come back. Didn’t have time.)
I’m 68 years old. I used to be able to hold a conversation on the phone and read a book at the same time, about something completely different. That disappeared when I was in my late 40s. The test requires, for the more difficult problems, testing many different hypotheses, if a clear pattern doesn’t pop up immediately. It’s almost certain that this takes more time for me now than when I was younger.
This almost certainly impacts my communication skills, for starters.
I would be careful with the interpretation of your results.
It is very uncommon to loose 46 points even over a whole lifetime, given the assumption that nothing bad happened to your brain. Intelligence is one of, if not even the most stable personality trait known to psychology. That is why losing more than two standard deviations without any apparent reason apart from ageing should be treated as a less likely explanation than either of the following ones:
You were compared to the wrong age group. An IQ of 100 is defined as the mean score for your age group. So if you were compared to people in their 20-30′s that would easily explain the unfavorable result. Your test score needs to be compared to 68 year olds (or perhaps 65 to 70 year old people). It’s quite safe to say however, that you got slower compared to your younger self and other young people for that matter.
Excerpt:
But a rather curious situation occurs when we examine the scores of gifted students on these
various sets of norms. In 1960, a five-year-old achieving a mental age of 8.0 would have had an IQ
score of 165. In 1972, that same raw score only yielded an IQ of 153, a difference of 12 points.
Differences between the Stanford-Binet Revision IV, published in 1986, and the 1972 norms
appear to be at least 13.5 points in the moderately gifted range (Thorndike, Hagen & Sattler, 1986),
which would bring the same child’s score down below 140. This is a loss of one IQ point per year
from 1960 to 1986 for children in the gifted range. In this 26 year period, average students needed
to obtain only 8 more points to make up for the average gains in intelligence of the general
population, whereas gifted children needed to obtain over 25 more points to match previous scores
1 1⁄2 standard deviations of IQ. This seems like an unreasonable demand.
I would be careful with the interpretation of your results.
Well, what I wrote was banter.
There are many kinds of intelligence. The test measures a particular kind, one that could probably be simulated (AI) with relative ease (I’m not saying it’s easy, but that what is involved is a series of tests, trials, of possible transforms, and then a checking of transforms that work for simplicity. It’s looking for an aha! pattern.
I know that I’m not as good at this now as I was when younger. A related example: I’m looking for my black waist pack, in my office, a room full of stuff. I walk through and don’t see it. We are in a hurry to leave, so I ask my 9-year-old daughter to check. She sees it immediately. It’s in plain sight. I have “tunnel vision.” Not literally. I still have peripheral vision. But I don’t interpret the full field, as I used to, only a narrower field, more central. I have to actually look at the bag to recognize it.
I trust the test as a reasonable one, that would measure a certain kind of intelligence that is highly useful.
Damn! I’m used to thinking of myself as really smart, for almost sixty years. Time to move on. Yes, I’m still smart in some ways, but I already knew that there are many ways in which I’m not, maybe never was.
What I’ve been told by doctors is that the cognitive impairments I’ve noticed are normal. People learn to compensate for them.
As age progresses, we also see a natural shift of intelligence from “fluid” to “crystallized” intelligence. The first kind is fast, adaptable and more creative, good for problem-solving, learning new things and pattern-recognition. The second kind is concerned with facts and knowledge, but also implicit knowledge/skills like how to drive a car.
IQ tests really measure fluid intelligence, less so the crystallized kind. Some IQ tests have a few questions that probe your crystallized intelligence as well, like “What was the name of the ship Charles Darwin sailed on to the Galapagos islands?” (often with 4 answers to choose from). But usually you get very few questions like those, if any at all.
Those two “kinds” of intelligence aren’t completely independent though, as one would expect your fluid intelligence has a high impact on your crystallized knowledge.
Interesting, Friendly-HI. I was pointing to something distinct from both. In the Wikipedia article, “crystallized intelligence” is not about “knowledge,” per se, but is something integrated. What has shifted for me is “fast,” when it comes to a series of new analyses of my sensory input. I’m not that kind of fast any more. However, “depth” appears to have increased.
To me, it’s important that I distinguish my accumulated experience from “truth.” It’s just my accumulated experience, my past. The present and future remain open, as long as I’m alive.
The test measures a particular kind, one that could probably be simulated (AI) with relative ease (I’m not saying it’s easy, but that what is involved is a series of tests, trials, of possible transforms, and then a checking of transforms that work for simplicity. It’s looking for an aha! pattern.
Here’s a 2010 Master’s Thesis that does pretty well on it. I remember someone came up with a better algorithm in the last year, but I’m not finding it quickly.
Friendly-HI, are you trying to explain Abd’s low score using the Flynn effect? Claiming that a modern IQ of 110 is equivalent to an IQ of 156 in the 1950s seems a bit like claiming that a modern skilled labourer is as clever as the people at Cal Tech in the 1950s. I suppose that’s possible, but I would be somewhat surprised. I mean, I haven’t noticed many people complaining that chess is too easy.
I’d imagine that it’s some combination of age-related decline and a badly calibrated test which has a large random-number generator component. (And a tiny bit of the Flynn effect)
Remember that this is the same scheme that puts a Richard Feynman in every 20 people while Marylin Vos Savant is an impossible genius who should not have occurred in the lifetime of the species.
Ideally your achieved IQ score is really a measure of your position within a normal distribution of IQ scores of your age group, where the mean (or peak) is standardized as 100 and one standard deviation equals 15 points. So an IQ of 130 is two Standard deviations above the mean and only ~ 2% of the people in your age group would be considered smarter than you.
I’m not sure age related decline factors into the decline of his IQ scores at all. That Hypothesis would only be true if the IQ-test he took was actually quite accurate and well-constructed, which would literally mean that in ~1955 only 0.05% of children in his age group were more intelligent than him and now something like 20% of ~65 year olds are more intelligent than him. Considering the stability of IQ it just doesn’t seem very plausible, that age-related decline would have hit him much harder than the average old person.
The article I quoted offered an explanation that I find much more plausible. It’s primary point wasn’t that this is the doing of the Flynn-effect, but the following:
“When too many children are found in the upper ranges, the scores are adjusted to fit the theoretical curve.This swells the number of scores in the 120-130 range and depresses the IQ scores of the entire gifted population. The attempt to artificially force the distribution of giftedness into the normal curve results in the disappearance of 1 1⁄2 standard deviations of intelligence. With today’s measuring devices, all IQ scores in the gifted range are most likely underestimates of ability.”
Whoa. I knew there was some weird stuff happening with high-IQ scores, but never realized it was this much.
Then again, most institutions I’ve seen that administer IQ tests seem to treat the formulas and scoring systems as corporate secrets or something. I should not be this surprised at the extent of the weirdness.
It is very uncommon to loose 46 points even over a whole lifetime, given the assumption that nothing bad happened to your brain.
You mean, nothing bad besides aging? If comparing Abd2012 to the correct age group would easily explain the numerical difference with regards to Abd1962′s score, then that’s solid evidence that this is regular age-related deterioration (of ability to solve Raven’s Progressive Matrices).
Given the same difficulty and grading of tests, then yes. However, there’s also the possibility that you have the same raw score on a test of the same level of difficulty, but achieve lower “IQ” once weighted and graded because of curious maths and changes in the base distribution model they use for grading.
I might be confusing what your point was though, on second thought. Am I?
So, it’s possible that a raw score one year will mean a different thing another year. For the SAT and GRE, getting one question wrong on the math section will drop you tens of points- but how many varies from year to year. (Other scores are more stable; that one is corrupted by edge effects of the tremendous number of people who get all the quantitative questions correct.)
The point I was making is that, when IQ is calculated by age group, that’s evidence that there are raw score differentials between age groups. This paper shows a theoretical graph of what that would like in Figure 1. Also related is Figure 3, but it has a crazy axis and so I’m hesitant to apply it. (I’m having trouble finding actual raw score data out there.)
If age-related decline and death are unrelated to intelligence, then even though raw scores will decline with age, individual IQ will stay the same in expectation (beyond unavoidable random drift) because each person is compared to people whose scores have declined about as much as theirs.
When IQ is used as a measure of “where are you relative to your peers?”, you want this. When IQ is used as a measure of absolute intelligence, you don’t want this. This email by Eliezer comes to mind.
“The point I was making is that, when IQ is calculated by age group, that’s evidence that there are raw score differentials between age groups.”
Exactly, that is the point. Of course there is a certain age-related deterioration of intelligence, especially fluid intelligence. So even if he did the exact same test he already did decades ago, his raw score will surely be lower now than it was back then. Confusingly enough, he could still be said to be as “intelligent” as he was back then if his relative position within the IQ distribution hadn’t changed. (Which if we were to believe his recent IQ-test, actually happened).
If any of this is confusing it’s because IQ is a relative measurement. So if I were to say that he is as intelligent as he was decades ago in the context of an IQ test, that doesn’t mean that he would solve the same proportion of tasks correctly, or that there wasn’t any cognitive decline due to aging, but only that his relative position within the normal distribution of IQ scores hasn’t changed.
IQ tests never measure absolute intelligence. Since IQ means intelligence -quotient-, you always compare a score to other scores, so it’s not an absolute measure by definition—there is no absolute IQ test. I’m also not aware of any respectable existing test for absolute intelligence either, nor how exactly one might even look like, although I’m sure you could in principle construct one if you define the word intelligence in nonconfused terms that reflect actual reality, which seems like a monumental task.
If we picture the concept of absolute intelligence as some kind of optimal information process with certain well defined characteristics whose lower and upper bounds are only determined by the laws of physics, I’m afraid human intelligence will be hardly comparable to it in any really meaningful way. And more importantly, how could you even begin to make a reliable and valid measure of something like that in humans?
Since IQ means intelligence -quotient-, you always compare a score to other scores, so it’s not an absolute measure by definition—there is no absolute IQ test. I’m also not aware of any respectable existing test for absolute intelligence either,
Right. Unfortunately, whenever someone wants to talk about absolute intelligence, “IQ” is the closest word/concept to that.
When you look at adult IQ tests, the raw score is decent measure of ‘absolute intelligence’ for most modern humans. Current tests have known problems with exceptional individuals (on either end) and some tests are more interested in determining the shape of someone’s intelligence (like, say, the subtests on the Woodcock Johnson) than others (like the Raven’s test, which only tests one thing). Comparing raw scores tells you useful things- about the effects of age, about the Flynn effect, about theoretical populations, and even about the distribution now. IQ scores are defined to follow a bell curve, but if the raw scores don’t follow a bell curve, that’s important to know!
The concept of IQ as a quotient seems rooted in the history of testing children- “this 12 year old has a 16 year old’s development”- which isn’t very useful for adults. If we give a test for adults to Alice and Betty, and Alice has an IQ of 140 and Betty has an IQ of 100, that doesn’t mean Alice is 40% smarter than Betty; it means that Betty is 50th percentile and Alice is 99.6th percentile. But, in practice, we might want to know that it takes Betty 90 seconds to get a problem right 80% of the time, and it takes Alice 5 seconds to get it right 100% of the time, which is data we collected in order to get the official outputs of 140 and 100.
If we picture the concept of absolute intelligence as some kind of optimal information process with certain well defined characteristics whose lower and upper bounds are only determined by the laws of physics, I’m afraid human intelligence will be hardly comparable to it in any really meaningful way.
The Sentience Quotient is the closest thing I can think of, and it’s mostly good for describing why humans and trees have few productive conversations (though the upper bound is also interesting).
There is little doubt in my mind that there is an age-related shift. Calling it “bad” would be shallow. There is a trade-off.
I don’t see it as a difference in “ability to solve,” but rather as a difference in the speed with which untrained heuristics can be used. That could be related to the effect I’ve long noticed, a marked decline in an ability to multiprocess, to handle multiple independent threads or processes. If solving the matrix involves testing a large number of possibilities, the more that can be tested at once, the faster the process will be. It’s as if I’ve moved toward being a Turing machine, from being massively parallel.
I would not consciously perceive the “separate processes,” necessarily. Rather, the result of them would pop up in my consciousness as “ideas.” I’d just “see” the solution.
The decline might be the result of increased capacity being devoted to depth rather than breadth. If so, it’s not a “bad” happening, but a relative disability related to an improvement in a different ability.
It points to certain issues in life extension, however. The brain might naturally reach a kind of saturation. Life extension without intelligence enhancement in some way, i.e., the development of cyborg technology, might not be all so valuable. (We are experiencing this to a degree in that we have rapid access to massive information, but the bandwidth of those connections is generally narrow.)
But these are just ideas. I have no specific test of “depth.”
Calling it “bad” would be shallow. There is a trade-off.
It certainly beats the alternative!
I would not consciously perceive the “separate processes,” necessarily. Rather, the result of them would pop up in my consciousness as “ideas.” I’d just “see” the solution.
Hm. When I originally read your description of solving the matrices, it seemed to me like your algorithm was shaped the wrong way- I would look at the matrix, identify the transformation, predict what the right answer would be, and then find it in the options. (I only used serious thought and hypothesis falsification on the last question.) Now I’m less confident that I understand my algorithm for identifying the transformation.
Hm. When I originally read your description of solving the matrices, it seemed to me like your algorithm was shaped the wrong way- I would look at the matrix, identify the transformation, predict what the right answer would be, and then find it in the options. (I only used serious thought and hypothesis falsification on the last question.) Now I’m less confident that I understand my algorithm for identifying the transformation.
That loss of confidence is a clue that you are understanding the process better.
How do you “identify the transformation”? That’s the whole banana!
There is a separate step, finding the answer in the set of answers, which is a partial confirmation. If one is not certain of the entire transformation, but has identified aspects of it, possible elements of the transformation, sometimes the choice can be made by elimination among the answers. But the process you describe is my own default, and that’s how I started. At first it was trivial. It got less simple. Then I saw that I was going to run out of time! Then it became a matter of optimizing what I was going to answer, once I got that I was unlikely to complete.
Obviously, I could take the test again, but that would defeat the purpose. I did go back to review certain problems, for the discussion here. Yes, to be a more standard intelligence test, the results should be reported by age. I suspect that, unless someone has trained for this kind of test, raw results will peak at a certain age, then decline after that.
Or the test could be untimed, in which case I’d expect I could do very well. I might do better than some younger people, just as “smart,” who aren’t as careful. I would not generally be satisfied with less than total, accurate prediction, with a simple algorithm. (Any answer could be justified with a complicated enough algorithm.)
Back to the question of how the transformation is identified. It’s an excellent question. It is questions like this that must be answered to develop artificial intelligence.
And for general artificial intelligence, they must be answered in the general case. It may be possible to find specific, “trick” algorithms that work for specific problems. But humans can solve these problems “out of the box,” so to speak, without almost no instruction. How do we do that?
Rather obviously, we are designed to detect patterns of behavior, which we use for prediction.
I wouldn’t be humbled just yet, especially if you found some of the problems impossible rather than complicated but doable. A lot of people seem to have got unexpectedly low scores on this test. (And no-one’s said ‘Wow I usually do really badly on IQ tests but that one gave me a great score’)
Then go and work out exactly what and why the answers on that test are as they are. (Perla has missed some of the explanations, but it’s very satisfying to work out what the answers actually are. They’re all perfectly logical and obvious-in-retrospect).
After that you should be reasonably confident that you’ll do very well on any similar tests in future.
What that tells you about the nature of IQ tests and their calibration is debatable.
Well, when I went back and looked at a couple of problems, I was able to solve them, so far. It was definitely, then, an issue of time. (When I find the solution, I expect, it is completely clear and the missing frame is fully specified, and it’s reasonably simple. I.e., “obvious in retrospect,” as you wrote.)
I do know, independently, that my “multiprocessing” abilities have declined, and that these would be likely be important to any algorithm for solving these problems. I’m sure I could improve my time with practice.
Thanks for your kind thoughts and for the link. I’ll check it out.
Took the IQ test. Humbling. Score 110.
IQ test in high school, 156. SAT 793⁄800 verbal, 783⁄800 math. Cal Tech. Yatta yatta. But that was many years ago. It’s pretty obvious what happened. Timed test. I only finished, in the time, about 2⁄3 − 3⁄4 of the questions, maybe a bit more, I didn’t keep count. (I skipped questions that weren’t popping up right away, thinking I’d come back. Didn’t have time.)
I’m 68 years old. I used to be able to hold a conversation on the phone and read a book at the same time, about something completely different. That disappeared when I was in my late 40s. The test requires, for the more difficult problems, testing many different hypotheses, if a clear pattern doesn’t pop up immediately. It’s almost certain that this takes more time for me now than when I was younger.
This almost certainly impacts my communication skills, for starters.
I would be careful with the interpretation of your results.
It is very uncommon to loose 46 points even over a whole lifetime, given the assumption that nothing bad happened to your brain. Intelligence is one of, if not even the most stable personality trait known to psychology. That is why losing more than two standard deviations without any apparent reason apart from ageing should be treated as a less likely explanation than either of the following ones:
You were compared to the wrong age group. An IQ of 100 is defined as the mean score for your age group. So if you were compared to people in their 20-30′s that would easily explain the unfavorable result. Your test score needs to be compared to 68 year olds (or perhaps 65 to 70 year old people). It’s quite safe to say however, that you got slower compared to your younger self and other young people for that matter.
Here is another explanation that may fit very nicely to your score pattern. http://www.gifteddevelopment.com/PDF_files/a35.pdf
Excerpt: But a rather curious situation occurs when we examine the scores of gifted students on these various sets of norms. In 1960, a five-year-old achieving a mental age of 8.0 would have had an IQ score of 165. In 1972, that same raw score only yielded an IQ of 153, a difference of 12 points. Differences between the Stanford-Binet Revision IV, published in 1986, and the 1972 norms appear to be at least 13.5 points in the moderately gifted range (Thorndike, Hagen & Sattler, 1986), which would bring the same child’s score down below 140. This is a loss of one IQ point per year from 1960 to 1986 for children in the gifted range. In this 26 year period, average students needed to obtain only 8 more points to make up for the average gains in intelligence of the general population, whereas gifted children needed to obtain over 25 more points to match previous scores
1 1⁄2 standard deviations of IQ. This seems like an unreasonable demand.
Well, what I wrote was banter.
There are many kinds of intelligence. The test measures a particular kind, one that could probably be simulated (AI) with relative ease (I’m not saying it’s easy, but that what is involved is a series of tests, trials, of possible transforms, and then a checking of transforms that work for simplicity. It’s looking for an aha! pattern.
I know that I’m not as good at this now as I was when younger. A related example: I’m looking for my black waist pack, in my office, a room full of stuff. I walk through and don’t see it. We are in a hurry to leave, so I ask my 9-year-old daughter to check. She sees it immediately. It’s in plain sight. I have “tunnel vision.” Not literally. I still have peripheral vision. But I don’t interpret the full field, as I used to, only a narrower field, more central. I have to actually look at the bag to recognize it.
I trust the test as a reasonable one, that would measure a certain kind of intelligence that is highly useful.
Damn! I’m used to thinking of myself as really smart, for almost sixty years. Time to move on. Yes, I’m still smart in some ways, but I already knew that there are many ways in which I’m not, maybe never was.
What I’ve been told by doctors is that the cognitive impairments I’ve noticed are normal. People learn to compensate for them.
As age progresses, we also see a natural shift of intelligence from “fluid” to “crystallized” intelligence. The first kind is fast, adaptable and more creative, good for problem-solving, learning new things and pattern-recognition. The second kind is concerned with facts and knowledge, but also implicit knowledge/skills like how to drive a car.
IQ tests really measure fluid intelligence, less so the crystallized kind. Some IQ tests have a few questions that probe your crystallized intelligence as well, like “What was the name of the ship Charles Darwin sailed on to the Galapagos islands?” (often with 4 answers to choose from). But usually you get very few questions like those, if any at all.
Those two “kinds” of intelligence aren’t completely independent though, as one would expect your fluid intelligence has a high impact on your crystallized knowledge.
http://en.wikipedia.org/wiki/Crystallized_intelligence
Interesting, Friendly-HI. I was pointing to something distinct from both. In the Wikipedia article, “crystallized intelligence” is not about “knowledge,” per se, but is something integrated. What has shifted for me is “fast,” when it comes to a series of new analyses of my sensory input. I’m not that kind of fast any more. However, “depth” appears to have increased.
To me, it’s important that I distinguish my accumulated experience from “truth.” It’s just my accumulated experience, my past. The present and future remain open, as long as I’m alive.
Here’s a 2010 Master’s Thesis that does pretty well on it. I remember someone came up with a better algorithm in the last year, but I’m not finding it quickly.
Friendly-HI, are you trying to explain Abd’s low score using the Flynn effect? Claiming that a modern IQ of 110 is equivalent to an IQ of 156 in the 1950s seems a bit like claiming that a modern skilled labourer is as clever as the people at Cal Tech in the 1950s. I suppose that’s possible, but I would be somewhat surprised. I mean, I haven’t noticed many people complaining that chess is too easy.
I’d imagine that it’s some combination of age-related decline and a badly calibrated test which has a large random-number generator component. (And a tiny bit of the Flynn effect)
Remember that this is the same scheme that puts a Richard Feynman in every 20 people while Marylin Vos Savant is an impossible genius who should not have occurred in the lifetime of the species.
Ideally your achieved IQ score is really a measure of your position within a normal distribution of IQ scores of your age group, where the mean (or peak) is standardized as 100 and one standard deviation equals 15 points. So an IQ of 130 is two Standard deviations above the mean and only ~ 2% of the people in your age group would be considered smarter than you.
I’m not sure age related decline factors into the decline of his IQ scores at all. That Hypothesis would only be true if the IQ-test he took was actually quite accurate and well-constructed, which would literally mean that in ~1955 only 0.05% of children in his age group were more intelligent than him and now something like 20% of ~65 year olds are more intelligent than him. Considering the stability of IQ it just doesn’t seem very plausible, that age-related decline would have hit him much harder than the average old person.
The article I quoted offered an explanation that I find much more plausible. It’s primary point wasn’t that this is the doing of the Flynn-effect, but the following:
“When too many children are found in the upper ranges, the scores are adjusted to fit the theoretical curve.This swells the number of scores in the 120-130 range and depresses the IQ scores of the entire gifted population. The attempt to artificially force the distribution of giftedness into the normal curve results in the disappearance of 1 1⁄2 standard deviations of intelligence. With today’s measuring devices, all IQ scores in the gifted range are most likely underestimates of ability.”
Whoa. I knew there was some weird stuff happening with high-IQ scores, but never realized it was this much.
Then again, most institutions I’ve seen that administer IQ tests seem to treat the formulas and scoring systems as corporate secrets or something. I should not be this surprised at the extent of the weirdness.
You mean, nothing bad besides aging? If comparing Abd2012 to the correct age group would easily explain the numerical difference with regards to Abd1962′s score, then that’s solid evidence that this is regular age-related deterioration (of ability to solve Raven’s Progressive Matrices).
Given the same difficulty and grading of tests, then yes. However, there’s also the possibility that you have the same raw score on a test of the same level of difficulty, but achieve lower “IQ” once weighted and graded because of curious maths and changes in the base distribution model they use for grading.
I might be confusing what your point was though, on second thought. Am I?
So, it’s possible that a raw score one year will mean a different thing another year. For the SAT and GRE, getting one question wrong on the math section will drop you tens of points- but how many varies from year to year. (Other scores are more stable; that one is corrupted by edge effects of the tremendous number of people who get all the quantitative questions correct.)
The point I was making is that, when IQ is calculated by age group, that’s evidence that there are raw score differentials between age groups. This paper shows a theoretical graph of what that would like in Figure 1. Also related is Figure 3, but it has a crazy axis and so I’m hesitant to apply it. (I’m having trouble finding actual raw score data out there.)
If age-related decline and death are unrelated to intelligence, then even though raw scores will decline with age, individual IQ will stay the same in expectation (beyond unavoidable random drift) because each person is compared to people whose scores have declined about as much as theirs.
When IQ is used as a measure of “where are you relative to your peers?”, you want this. When IQ is used as a measure of absolute intelligence, you don’t want this. This email by Eliezer comes to mind.
“The point I was making is that, when IQ is calculated by age group, that’s evidence that there are raw score differentials between age groups.”
Exactly, that is the point. Of course there is a certain age-related deterioration of intelligence, especially fluid intelligence. So even if he did the exact same test he already did decades ago, his raw score will surely be lower now than it was back then. Confusingly enough, he could still be said to be as “intelligent” as he was back then if his relative position within the IQ distribution hadn’t changed. (Which if we were to believe his recent IQ-test, actually happened).
If any of this is confusing it’s because IQ is a relative measurement. So if I were to say that he is as intelligent as he was decades ago in the context of an IQ test, that doesn’t mean that he would solve the same proportion of tasks correctly, or that there wasn’t any cognitive decline due to aging, but only that his relative position within the normal distribution of IQ scores hasn’t changed.
IQ tests never measure absolute intelligence. Since IQ means intelligence -quotient-, you always compare a score to other scores, so it’s not an absolute measure by definition—there is no absolute IQ test. I’m also not aware of any respectable existing test for absolute intelligence either, nor how exactly one might even look like, although I’m sure you could in principle construct one if you define the word intelligence in nonconfused terms that reflect actual reality, which seems like a monumental task.
If we picture the concept of absolute intelligence as some kind of optimal information process with certain well defined characteristics whose lower and upper bounds are only determined by the laws of physics, I’m afraid human intelligence will be hardly comparable to it in any really meaningful way. And more importantly, how could you even begin to make a reliable and valid measure of something like that in humans?
Right. Unfortunately, whenever someone wants to talk about absolute intelligence, “IQ” is the closest word/concept to that.
When you look at adult IQ tests, the raw score is decent measure of ‘absolute intelligence’ for most modern humans. Current tests have known problems with exceptional individuals (on either end) and some tests are more interested in determining the shape of someone’s intelligence (like, say, the subtests on the Woodcock Johnson) than others (like the Raven’s test, which only tests one thing). Comparing raw scores tells you useful things- about the effects of age, about the Flynn effect, about theoretical populations, and even about the distribution now. IQ scores are defined to follow a bell curve, but if the raw scores don’t follow a bell curve, that’s important to know!
The concept of IQ as a quotient seems rooted in the history of testing children- “this 12 year old has a 16 year old’s development”- which isn’t very useful for adults. If we give a test for adults to Alice and Betty, and Alice has an IQ of 140 and Betty has an IQ of 100, that doesn’t mean Alice is 40% smarter than Betty; it means that Betty is 50th percentile and Alice is 99.6th percentile. But, in practice, we might want to know that it takes Betty 90 seconds to get a problem right 80% of the time, and it takes Alice 5 seconds to get it right 100% of the time, which is data we collected in order to get the official outputs of 140 and 100.
The Sentience Quotient is the closest thing I can think of, and it’s mostly good for describing why humans and trees have few productive conversations (though the upper bound is also interesting).
There is little doubt in my mind that there is an age-related shift. Calling it “bad” would be shallow. There is a trade-off.
I don’t see it as a difference in “ability to solve,” but rather as a difference in the speed with which untrained heuristics can be used. That could be related to the effect I’ve long noticed, a marked decline in an ability to multiprocess, to handle multiple independent threads or processes. If solving the matrix involves testing a large number of possibilities, the more that can be tested at once, the faster the process will be. It’s as if I’ve moved toward being a Turing machine, from being massively parallel.
I would not consciously perceive the “separate processes,” necessarily. Rather, the result of them would pop up in my consciousness as “ideas.” I’d just “see” the solution.
The decline might be the result of increased capacity being devoted to depth rather than breadth. If so, it’s not a “bad” happening, but a relative disability related to an improvement in a different ability.
It points to certain issues in life extension, however. The brain might naturally reach a kind of saturation. Life extension without intelligence enhancement in some way, i.e., the development of cyborg technology, might not be all so valuable. (We are experiencing this to a degree in that we have rapid access to massive information, but the bandwidth of those connections is generally narrow.)
But these are just ideas. I have no specific test of “depth.”
It certainly beats the alternative!
Hm. When I originally read your description of solving the matrices, it seemed to me like your algorithm was shaped the wrong way- I would look at the matrix, identify the transformation, predict what the right answer would be, and then find it in the options. (I only used serious thought and hypothesis falsification on the last question.) Now I’m less confident that I understand my algorithm for identifying the transformation.
That loss of confidence is a clue that you are understanding the process better.
How do you “identify the transformation”? That’s the whole banana!
There is a separate step, finding the answer in the set of answers, which is a partial confirmation. If one is not certain of the entire transformation, but has identified aspects of it, possible elements of the transformation, sometimes the choice can be made by elimination among the answers. But the process you describe is my own default, and that’s how I started. At first it was trivial. It got less simple. Then I saw that I was going to run out of time! Then it became a matter of optimizing what I was going to answer, once I got that I was unlikely to complete.
Obviously, I could take the test again, but that would defeat the purpose. I did go back to review certain problems, for the discussion here. Yes, to be a more standard intelligence test, the results should be reported by age. I suspect that, unless someone has trained for this kind of test, raw results will peak at a certain age, then decline after that.
Or the test could be untimed, in which case I’d expect I could do very well. I might do better than some younger people, just as “smart,” who aren’t as careful. I would not generally be satisfied with less than total, accurate prediction, with a simple algorithm. (Any answer could be justified with a complicated enough algorithm.)
Back to the question of how the transformation is identified. It’s an excellent question. It is questions like this that must be answered to develop artificial intelligence.
And for general artificial intelligence, they must be answered in the general case. It may be possible to find specific, “trick” algorithms that work for specific problems. But humans can solve these problems “out of the box,” so to speak, without almost no instruction. How do we do that?
Rather obviously, we are designed to detect patterns of behavior, which we use for prediction.
I wouldn’t be humbled just yet, especially if you found some of the problems impossible rather than complicated but doable. A lot of people seem to have got unexpectedly low scores on this test. (And no-one’s said ‘Wow I usually do really badly on IQ tests but that one gave me a great score’)
Go and read http://www.jperla.com/blog/post/how-to-ace-an-iq-test (don’t look at the answers, just the methods) and then go back and redo the test. I imagine you’ll then get a much higher score.
Then go and work out exactly what and why the answers on that test are as they are. (Perla has missed some of the explanations, but it’s very satisfying to work out what the answers actually are. They’re all perfectly logical and obvious-in-retrospect).
After that you should be reasonably confident that you’ll do very well on any similar tests in future.
What that tells you about the nature of IQ tests and their calibration is debatable.
The URL is incorrect, the comma at the end should be removed. Here is the page
Well, when I went back and looked at a couple of problems, I was able to solve them, so far. It was definitely, then, an issue of time. (When I find the solution, I expect, it is completely clear and the missing frame is fully specified, and it’s reasonably simple. I.e., “obvious in retrospect,” as you wrote.)
I do know, independently, that my “multiprocessing” abilities have declined, and that these would be likely be important to any algorithm for solving these problems. I’m sure I could improve my time with practice.
Thanks for your kind thoughts and for the link. I’ll check it out.