Do you trust IQ tests?

[Tsuyoiko]

I took the Mensa test as teenager and scored 148. In online tests I usually score around 150.

 

I think IQ tests do test something real, i.e., how good you are at completing IQ tests. I think intelligence is a lot more complex than the results of one type of test can measure.

[John Cuthber]

I don't know a lot about IQ testing but I do know about puns.

http://xkcd.com/559/

[jake.com]

Well, that was the IQ test. Your response corresponds to an IQ of 71.

 

darn, well maybe next time

[lakmilis]

tsuyoiko.. if you did the paper version of a mensa test and scored 148.. you would score 258 on online ones.. I swear!!!

 

well coke.. so we both suddenly became trolls.... let's go and change the world now, shall we?

 

Ye.. Good thing what you say about the bubble.. just wasn't sure.

[coke]

lakmilis are you ever going to tell me and jake what pun was not intended?

 

come on, we're dying to know what pun was not intended...

 

LOL... 'NO'I don't think facebook iq tests are likely to be 'accurate'

 

(no pun intended)

 

what pun not intended?

 

come on... soemone tell me what pun was not intended...

I don't understand either.

 

Was this the pun you didn't intend?

I don't know a lot about IQ testing but I do know about puns.

http://xkcd.com/559/

And John' date=' I don't get what pun was not intended in that comic either!

 

Here's a bad pun from one of the threads...

 

I'm dating a chemist, something something what do I do?

You'll just have to wait and see if there's chemistry between you.

 

Hahaha hilarious. Let's make this thread about puns.

Edited April 7, 2009 by coke

[lakmilis]

lol well coke two things.. that pun comic strip made me laugh... if you don't get it... well, that's kind of the point of it... the guy pondering afterwards.. due to randomness... like eg. my 'no pun intended'.

 

Here's your answer... check the date of the post

 

(any other day but april fool's day... perhaps... 'no offence' would be the actual term where 'no pun intended' was used).

 

lak

 

hope you are happy cocaine or coca cola

[lakmilis]

 

Just so you know, clicking all the true answers as fast as you can on this test gives you an IQ of 118. Also, these questions do not measure only intelligence, but also knowledge. Also, a multiple choice test must subtract a portion of the points (equal to [math]1/(n-1) times the questions' value where n is the number of choices) for a wrong answer if the test is to be worth anything.

 

 

Hi.. I just spoke to someone who was really is?/was really important to me; I mentioned something and I came across the scienceforums again and I see this post.. and LOL .. now usually I think I can rememeber that skeptic is a reasonable person.. but wait... if one answers a test (regardless of its validity), at the fastest possible time skeptic claims, (a universal highest speed from nerves, muscles , and thought), one gets 118 , when one chooses all the right answers... so just how did we get like loads of various higher scores? by doing the wrong ones???

[PhDwannabe]

Hi; this seems to be a fairly old post, but I just registered here, and I thought I'd weigh in. I'm a PhD student in clinical psychology, and have done some amount of testing. I'll see if I can make some illustrative comments on the matter.

 

First and foremost, the biggest confusion I tend to see in discussions like this--present one not excepted--is the matter of what "intelligence" really is. We often hear people saying "he does well on IQ tests, but he isn't that intelligent," or even further removed, "isn't that smart." Sometimes, comparisons between intelligence and common sense, or "book smarts and street smarts" are made. Very often, critique of intelligence testing tends to come from a distrust of egghead scientists and their goofy games. (If you think that public perception of psychologists is anywhere but in the toilet, ask yourself how many psychologist movie or TV characters ever 1) are good guys or 2) get anything right.) The biggest thing I can say is that most of us are not so proud of our instruments as to regard them as the final say on a person or their functioning. In short, I and most of my colleagues believe that a score on an IQ test can be a really important part of the picture--deliberate emphasis on part.

 

So, what is it, exactly? Well, depends on who you ask. My answer is a fairly middle-of-the-road one. Many people even outside of my field know that the first widely-adopted measurement of cognitive ability was one developed by the French psychologist Binet, on a commission from his government to improve services for schoolchildren we would today call developmentally disabled. Lewis Terman of Stanford eventually did some work on it, thus creating the Stanford-Binet, which some older forum members here may even remember taking in school. (Though it's still around in some form, the most commonly used tests are now the Wechslers: the WAIS-IV for adults and the WISC-IV for children.) In these early days, Charles Spearman, one of psychology's granddads, noted that when we give separate kinds of tests for intelligence--like, for instance, quickly putting complicated block designs together, or solving logic problems--the scores, though different for these different skills, all seemed to be related to one another pretty strongly. His idea was that there was some big factor underlying all of these tests, with smaller factors on top of it adjusting the scores a bit up and down for different skills--the big, underlying factor he called g, for "general intelligence." Though we've moved on quite a bit from Spearman's ideas, we still use the idea of g in a somewhat similar way. IQ is the number our tests give us; g is the hidden construct that we can't really stick a ruler next to and directly measure. Our aim is to approximate measurement of g as closely as possible. Most of us are perfectly happy to admit this.

 

So, what's g, then? What is this thing we call intelligence? Again, depends on who you ask. We can't see it or touch it or taste it, so we define it somewhat arbitrarily. (One is reminded of Justice Potter Stewart's famous observation about pornography--that he couldn't define it, but he knew it when he saw it.) David Wechsler's oft-quoted phrase is: "The aggregate or global capacity of the individual to act purposefully, to think rationally, and to deal effectively with his environment." It's the ability of your reasoning tools to direct adaptation to an environment, or to the general sort of environment some creature like you might expect. (When you look at it this way, you see why species-to-species IQ comparisons are so stupid. A raccoon is a hell of a lot smarter than us for the purposes of raccoon life and livelihood.) This construct isn't real in the usual sense of the word (I often tell my students: "A little more real than a unicorn, a little less real than the chalkboard.") It derives most of its meaning from its relationship to other things which seem to exist more solidly in the real world: academic achievement and occupational attainment are sort of the big two. A supposedly strong measure of g is worthless if it doesn't predict well the sorts of things we expect intelligence to be related to, like doing well in school and performing well in a job. Well, with only a century of work behind us, I can tell you that our measures actually do do that, and fairly well. Beyond this, they're also good for identifying specific deficits, like impairments in reading ability, spatial reasoning, etc. (Tommy's mere poor performance on his homework doesn't really tell you what's driving the problem.) An IQ test is often a good first step to seeing some of these deficits in a child who is struggling, and point the way to more specific cognitive testing or other services. I could go on with other useful functions of the tests, but I think you get the idea.

 

So, what is intelligence not? The brain-as-computer analogy has a million problems with it, but we'll use it here for a moment anyway. A high IQ might be thought of as a really fast processor and a great deal of RAM. A nice machine, sure, but nothing about its niceness prevents you from installing terrible software on it. A high IQ is not particularly protective against mental illness or unhappiness, nor does it prevent a person from becoming (to steal a phrase from this thread!) a "maladjusted loser." I swear, most of us aren't as obsessed with the results of our shiny tests as we are often perceived to be--if you sat down and talked to a psychologist, many would happily tell you that an IQ test only tells you about a small piece of a person. It's moderately correlated with what might be outwardly observable as "smarts," because accumulated experience and knowledge (we sometimes call this "crystallized intelligence") matter so much more than anything else in so many situations. If I were crash-landed in the middle of the Alaskan wilderness, I'd want the 90 IQ experienced bush pilot over the 140 IQ accountant to survive with me! Of course, if I got to pick between the 90 bush pilot and the 140 bush pilot, well, that's an easy one too.

 

All that said, it really is good for something and means something--just not everything. If you know your own IQ, don't expect it to be amazingly predictive of your life trajectory. These things are predictive of important variables we see in populations. That doesn't necessarily translate to every individual, since we are creatures of such substantial variability. That doesn't mean the variables or the constructs don't have a great deal of meaning (Not everyone who drinks and is suicidal kills themselves, but I'm a hell of a lot more worried about a suicidal patient who drinks, because the rates are so much higher there than among those who don't drink.)

 

Oh, and finally, the writing, development, and validation of tests like these is like building a damned moon rocket. An unbelievable amount of extremely complicated work. One should not expect any half-assed internet version to say anything reasonable. And to those of you who have taken an internet version and found the score very close to that from a real, validated instrument? Well, buddy, every blind squirrel finds a couple of nuts.

 

 

Thanks,

DJ

[Mr Skeptic]

Hi.. I just spoke to someone who was really is?/was really important to me; I mentioned something and I came across the scienceforums again and I see this post.. and LOL .. now usually I think I can rememeber that skeptic is a reasonable person.. but wait... if one answers a test (regardless of its validity), at the fastest possible time skeptic claims, (a universal highest speed from nerves, muscles , and thought), one gets 118 , when one chooses all the right answers... so just how did we get like loads of various higher scores? by doing the wrong ones???

 

Yes, you get a very high score on some tests simply for taking the test. Right answer, wrong answer, what's the difference? Especially the "IQ tests" that want to make money by giving you a full report on how smart you are. If you want a reliable IQ test look for one that isn't trying to make money over you, or at least one that you have to pay ahead of time rather than them trying to bribe you into buying something by telling you how smart you are.

[needimprovement]

Do you think IQ tests are a useful way to measure intelligence; or do you think they're useful at all?

IQ test are a very accurate predictor - of how well you do on IQ tests! Otherwise, not very useful.

[PhDwannabe]

IQ test are a very accurate predictor - of how well you do on IQ tests! Otherwise, not very useful.

 

Do you have any kind of peer-reviewed empirical data to back this up? I can't wait to see it.

[wright496]

I think IQ tests are a sort of decathlon of mental abilities. We have a much better understanding of neurology than when these tests were first devised. I believe we should be precise about what we are testing for, such as working memory, attention, etc. The best support of these tests seems to come from the fact that they correlate with so many things. IQ tests take on so much at once though, that it's hard to elucidate what's really going on. Let's say hypothetically (and I'm just pulling this out of my ass to make a point) that verbal skills make a big difference in job performance. Clearly any IQ test which took into account verbal skills would correlate with good verbal skill and so IQ and job performance would be correlated as well. Well, you could say IQ and job performance are related and leave it at that. But WHY is IQ related to job performance. Is it working memory, or attention or math skills? We need to break it down more and analyze these separate cognitive parameters and their interrelationships.

[needimprovement]

Do you have any kind of peer-reviewed empirical data to back this up? I can't wait to see it.

I was just referring to my personal experience (actually combined aptitude and intelligence) so I can’t show you any kind of peer-reviewed empirical data. Hence, it then does not prove that IQ tests are not useful.

 

In the Philippines IQ tests are being used in schools and workplaces to determine eligibility and placement for many years now. Maybe they also have some weakness and some IQ tests are bogus.

Edited August 12, 2010 by needimprovement

[Ringer]

http://www.amazon.co...s/dp/0465025102

 

good book on the subject IMO. I'm about halfway through it and it makes some good points.

 

http://www.indiana.e...ynneffect.shtml

 

also and interesting read and pretty good overview of IQ and the Flynn effect.

Edited August 11, 2010 by Ringer

[PhDwannabe]

I believe we should be precise about what we are testing for, such as working memory, attention, etc.

We are. IQ testing reports include more than a score. Clinicians describe the breakdowns of test subscales and the relationship between subscale scores, and what implications that has or may have for the individual. IQ tests are also often part of a testing battery which evaluates skills in an even more focused manner, or helps inform the choice of assessments for a future battery.

 

IQ tests take on so much at once though, that it's hard to elucidate what's really going on.

It's not. If you read an IQ report, rather than relying on a single FSIQ ("Full Scale Intelligence Quotient") score.

 

But WHY is IQ related to job performance. Is it working memory, or attention or math skills? We need to break it down more and analyze these separate cognitive parameters and their interrelationships.

We do. Again, that's really what an IQ test is for. To give both a broad measure of a person's general intellectual ability ("g"), as well as more comprehensive accounts of a person's specific capacities in several different areas. Of course, these areas are typically highly intercorrelated, which lends meaning to the concept of general intelligence. I wrote a relatively gigantic post about two weeks ago, right up the page there, that hit this and several other issues.

[Mr Skeptic]

PhDwannabe, you seem to be our local expert on IQ tests. I've noticed that most online IQ tests are absolute rubbish. Do you know any good ones?

[PhDwannabe]

PhDwannabe, you seem to be our local expert on IQ tests. I've noticed that most online IQ tests are absolute rubbish. Do you know any good ones?

 

I wish I did! They're fun, after all. No, I really don't though--if, as I've assumed, by "good" you mean "empirically validated to even a minimum extent." At present time, you really need a human to perform the test. I'll ring this bell as much as I can: the real result of an IQ test is not an FSIQ score, it's a narrative assessment report. The clinician-tester is also doing a fair amount of observation in addition to the plain old mechanics of the test, and that sort of trained/careful observation and interpretation of the test is really important to a full understanding of a person's intellectual capacities.

 

Oh yeah, and even if this wasn't the case, well, the tests cost so damn much to properly develop that you'd see a "real" one being given out for free on the internet right after you see your local pharmacy giving away medication. Sorry to disappoint!

Edited August 12, 2010 by PhDwannabe

[needimprovement]

I'll ring this bell as much as I can: the real result of an IQ test is not an FSIQ score, it's a narrative assessment report. The clinician-tester is also doing a fair amount of observation in addition to the plain old mechanics of the test, and that sort of trained/careful observation and interpretation of the test is really important to a full understanding of a person's intellectual capacities.

Yes - We should not be blinded by the results of an IQ test.

 

For competence we need: abilities, skills and knowledge.

 

Abilities come from recruiting or promoting the right people.

 

Skills and knowledge come from effective learning.

 

Training people with the wrong abilities is a waste of money.

 

IQ is part of the abilities but incomplete.

 

Psychometric testing determines how the person is wired and their natural abilities (aka gifts or talents).

[wright496]

I don't normally like to stray off topic, but since it's closely related and this thread has been stagnant anyway, what do you guys think of the Flynn Effect?

[Mr Skeptic]

I don't normally like to stray off topic, but since it's closely related and this thread has been stagnant anyway, what do you guys think of the Flynn Effect?

 

An increase of IQ of about 3 points per decade? I'll go with better nutrition and a richer environment. Taking a tangent on your tangent, do you think that the Flynn effect could be a basis for Technological Singularity?

[Dan6541]

Personally I think IQ tests are overrated.

 

I live in Victoria, Australia, and in this state the government is using IQ tests to determine whether children with aspergers and autism should attend a normal school or a "special" school. My understanding of these disorders is that they affect peoples social capabilities, not their general IQ. So, yes, maybe it is good for a rough estimate but it's being used for purposes that aren't proper considering what kind of test it is.

[Sisyphus]

Personally I think IQ tests are overrated.

 

I live in Victoria, Australia, and in this state the government is using IQ tests to determine whether children with aspergers and autism should attend a normal school or a "special" school. My understanding of these disorders is that they affect peoples social capabilities, not their general IQ. So, yes, maybe it is good for a rough estimate but it's being used for purposes that aren't proper considering what kind of test it is.

 

The autism spectrum is about a whole lot more than social skills. I think a test that measures to what degree on is capable of thinking in the needed ways sounds reasonable, since presumably that's what is at issue. I don't know how good a tool it actually is for that, but to me it's at least not obviously useless.

[Dan6541]

The autism spectrum is about a whole lot more than social skills. I think a test that measures to what degree on is capable of thinking in the needed ways sounds reasonable, since presumably that's what is at issue. I don't know how good a tool it actually is for that, but to me it's at least not obviously useless.

 

Sure autism is about more than about social skills but I think sending someone with a lack of these skills to a normal school purely because they got a normal score in an IQ test (a test that doesn't test social capabilities) is wrong.

[wisejanesm]

IQ tests should include different types of questions. Questions should be: with words, with pictures, with numbers, with letters, with three-dimensional images.

 

That's excellent IQ tests:

http://www.treningmozga.com/tests_iq_en.html

Such tests can be trusted.

[Wrynn]

I am not sure any test can actually estimate our intelligence because there's so much more to it. I remember seeing so called IQ Tests which were basically all mathematical. I understand that mathematics is queen of all the sciences but really? LIke someone said, maybe IQ and intelligence are not so all together. There are so many cultural dependancies too which need to be taken under consideration.