My point is that even in PE, there are different sports and different capabilities that people can display. Someone who can lift very heavy weights may not be a very fast runner, and vice versa.
As another grownup, I’d like to point out that not all jobs are assembly line jobs as in your example and instead require decision making in uncertain environments (i.e., there is no right answer). So if the deadline is on Friday, the deadline is on Friday… your answer will not be any more accurate if you get an extra two weeks… if at all, it’ll be completely useless by then.
OK, let me give you an example. Four grownups - A, B, C, and D - are given an open-ended problem with multiple solutions. A comes up with a very quick solution that is not very elegant, but works decently well. B takes more time and comes up with multiple solutions, all of which can work, and which have different advantages and disadvantages. C takes even longer and comes up with a simple and elegant solution that is much better than any of the options offered by A or B, and that neither A nor B could have come up with. D takes a medium-long amount of time but delivers a thorough analysis of the problem and a reason why it’s not even necessary to solve it because a way of avoiding the whole situation exists. Which one of them is the smartest? The IQ test will favor A.
Also, “reasoning” is not a single dimension. Reasoning about what? Processing what? Problem-solving in what category? We are not single-purpose machines. Are we just talking about logic problems on the LSAT? Or are we talking about Math Olympiad-type problems? Are we talking about linguistic puzzles? Are we talking about music theory problems? Problem-solving in the social realm? Are we talking about the ability to come up with creative solutions to open-ended problems? The ability to retain information well and recall it easily? This is what I was getting at - if all you’re measuring is the ability to solve logic puzzles as quickly as possible, you’re getting a measurement that’s pretty damn useless. And I speak as someone who is very good at solving logic problems as quickly as possible.
Just throwing in an idea about beings which are smarter/different than human IQ.
So the thing is, humans make IQ tests for other humans (or we make tests for animals). What about, say, an alien or being which has “higher than human” IQ. Would it make sense to make an IQ test for someone with greater IQ than the person who made it?
What if a being with higher IQ than me saw me make an IQ test?
What if I gave myself an IQ test?
Would it make sense to give myself an IQ test? After all, it’s my IQ.
What if I failed my own IQ test? Does that make sense?
Or what if I did better than my own IQ test, because I change over time?
Evidently this isn’t always the case though… for instance there are 7.9 billion people in the world and yet the distribution of our wealth is extremely positively skewed and isnt even close to following the normal curve.
The most prevalent and commonly given examples of variables which are normally distributed are ones which are assigned by nature. Like the 3 examples I gave above. Clearly not all variables which follow the normal curve are determined by nature, but…
The Normal Distribution occurs when many independent random factors affect an outcome
This is the main reason why I hypothesize that IQ scores won’t follow the normal curve, and would probably be heavy-tailed
There is nothing random going on in IQ tests, everything is well within the test-takers control.
If we asked every member of the population to pick a number 1-10, would the results be normally distributed with a mean of 5? of course not.
I think that this is the same reason why high school exam results (at least in my country) are also heavy-tailed. A large number of highly motivated people score between 90-100%, while less people score between 75-85%. If everyone in the world took an IQ test today, the distribution would most likely have a very heavy right tail. And I suspect that if we gave everyone 12 months “warning” before taking the test (like an exam), then it probably would bear no resemblance to the normal curve whatsoever.
I would like to add that the creator of the test has the easy part. I can create a sudoku, but after I have created it, you need to actually apply the techniques in order to solve it. Also the creation of the sudoku does not have to be done in a single day, I can spend months designing it in order to work the way I intended.
A similar argument once passed by on the Writing reddit page, asking how an averagely intelligent writer could write about a highly intelligent character. One of the best answers I came across also mentioned time. You as the writer can spend weeks writing what the character says in a 3-minute conversation.
The central limit theorem (CLT) states that the distribution of sample means approximates a normal distribution as the sample size gets larger, regardless of the population’s distribution.
I have no idea what it is that you are trying to explain here… do you know the difference between a sample and a population? I mean if you’re missing the relevant background in statistic it’s not really worth talking about it. And that’s fine, a lot of people never studied statistics.
Why don’t you just test your hypothesis then? Sample data is clearly out there and then you can either accept or reject your hypothesis. It’s a bit of a foregone conclusion though, seeing how you know that the average is 100 and the standard deviation is 15… the fact that it’s normalized kinda gives it away… it’s a normal distribution.
You’re talking about 90-100% on an exam out of 100%… you’d have to make one class take a test one year, normalize that result, and the next year have another class take the test and then give their scores according to the curve from the previous year… that’s how IQ scores are given.
Guys, I’m sorry… but you can’t come up with your own definitions of everything to question already defined terms. Reasoning and problem-solving abilities refers to fluid and crystallized intelligence. These are established terms in Psychology.
Here on the other hand, people argue that “well, if breakfast is me breaking my overnight fast… what if I had a cookie half an hour before, is breakfast still breakfast then?”
The original question (i.e., title of the post) was “Do IQ tests really measure intelligence” which I already said would be like asking “Does a GMAT tell you about your management skills?” or “Does the LSAT really tell you how good a lawyer you’ll make?”
Let me stop your right there… there is only one grownup because you can’t afford to hire four of them just to entertain this thought experiment; which is all this is because it has nothing to do with real life.
If B, C, and D don’t meet the deadline and A is the only one with a solution before the deadline, then there’s nothing wrong with that.
Well, it’s either an idea or what usually happens in any thread about IQ on this forum, first people start giving their own definitions of intelligence or what it should be and how it should be measured… and then things get philosophical. Same old, same old really…
…plus, what does smarter than human IQ even mean? Smarter than the mean of 100? I think here you’re actually referring to intelligence and not intelligence quotient (IQ)… just saying.
And this is how you end up with superficial thinking, Internet arguments that involve no in-depth analysis of anything, and engineered devices that break after 10 minutes of use because no one spent any time thinking about the matter for more than 5 minutes.
Fluid and crystallized intelligence are, in fact, established terms in psychology. Whether an IQ test is a good way to measure any of that is an open question. Whether these are useful concepts that actually measure anything useful in the real world is also an open question. “Established” doesn’t mean “useful”. Phrenology terms used to be established too.
I speak as someone who, way back when, made it a bit of a hobby of taking IQ tests. I took two officially proctored IQ tests in the same condition in the same situation. The results were 50 points apart. If I measured my temperature with two different thermometers and got results like that, would anyone consider those thermometers any good?
A quick google search reveals it’s been done a few times in independent studies. For example:
which found that the distribution had a particularly heavy right tail, as I had assumed
“Having become suspicious that the distribution of IQ scores at the high end appeared not to conform to normal curve expectations, I graphed the real data, from what I thought to be the most useful and reliable sources, against the normal curve. The results can be seen in the graph above. There are, indeed, a far higher number of people scoring at the high end than would be predicted by the normal curve.”
It doesn’t really matter though… performance on IQ tests is just a skill that can be learned. So it shouldn’t matter whether that skill is measured on a scale of 1 to 10, or 0 to a million, or with a mean of 100 like it is on IQ tests, it is still just a skill that isn’t going to be normally distributed because its been well established that very few skills are.
By normalized I assume you mean taking a Z score of the raw data. pretty sure all that does is allows them to compare the raw results against their hypothetical distribution (μ=100,σ=15), and doesn’t actually change the shape of the data, but maybe you can enlighten me on this.
But even so, if they have to standardize the data afterwards in order to get it to follow the normal distribution then doesn’t that just prove my point that our skill in taking IQ tests does not follow it?
I am willing to concede that…
“People tend to think of IQ as being normally distributed, but it is really just IQ scores that are normally distributed”
“And how do IQ scores get to be normally-distributed? The questions on the IQ tests get tweaked, added, and dropped so that the scores do not bunch too much at the low or high end” 3 Myths About the Normal Distribution | Elder Research
But it doesn’t change the fact that mankind’s ability to solve the puzzles on IQ tests should not be normally distributed.
Oh boy… I think all you’re doing here is give @meepster more anecdotal evidence that “quick” means “quick and dirty.” So, aside from the fact that this webpage should somewhere have a banner that reads best viewed in Netscape Navigator 2.0, let’s have a look at what you managed to find us there…
The title should really read “A commentary on … by abelard” because the way it stands, you are mislead into thinking that this is in fact the article by Cyril Burt, who lived from 1883-1971 and published the discussed report The Distribution and Relations of Educational Abilities in 1917 already.
Among some gifted education researchers, advocates, and practitioners, it is sometimes believed that there is a larger number of gifted people in the general population than would be predicted from a normal distribution (e.g., Gallagher, 2008; N. M. Robinson, Zigler, & Gallagher, 2000; Silverman, 1995, 2009), a belief that we termed the ?overabundance hypothesis.? We tested this hypothesis by searching public datasets and the published literature for large representative datasets, 10 of which were found in 6 sources. Results indicated that the overabundance hypothesis was mostly unsupported by the data. Moreover, most datasets included approximately the same (or fewer) gifted individuals than would be predicted from a normal distribution. We conclude the article by exploring the theoretical reasons why the overabundance is likely untrue and why some might believe otherwise.
Of course “quick” means “superficial”. It’s just a matter of common sense - thinking takes time. The more time you take, the more thinking you can do. The more time you take, the more reading you can do and the more analysis you can do of what you read.
And it is very much true that we prioritize quick-and-dirty over slow-and-good. Fast food, fast thinking - and the thinking is pretty much as good as the food, in a lot of cases.
