Are IQ tests still widely used today?

If they are, then what do people still use them for?

They were never “widely” used for some values of widely. Real IQ tests are expensive to administer and take several hours one-on-one with a highly trained professional. There isn’t much utility in giving them to anyone that falls anywhere in the normal range except to develop the test itself.

However, they are still used to identify people on both the high and low end of the scale. Many gifted and talented programs require them to be admitted. Societies like Mensa accept valid scores. Thy are also used to identify people with special needs or specific cognitive impairments. They are also sometimes used in research studies.

So, in high school 1990-1994, my psychology teacher said everyone in the US was tested for IQ, and once someone was 18 or over they could contact the Department of Education for the results.

I do remember some tests back in the day, like, determining the sequence of the stages of a broken light bulb, etc.

Is this true?

No. (Sadly, not much more to say in response than that.)

No. It would take forever and cost too much. I was once trained to administer several different kinds of IQ tests and it is an intensive process. Written tests like the SAT tend to correlate fairly highly with real IQ tests like the Stanford-Binet but they aren’t and most people have never taken a real one. It costs at least hundreds of dollars per person. There is no point to that to just find out that you are about average.

Within the last decade:

They’re still used by Michigan Rehabilitation Services (MRS) to evaluate what kind of jobs they should be steering people towards.

They’re also still used by Social Security to determine if someone has severe enough mental deficiencies to be considered automatically disabled.

At least, those are the reasons I assume I took them. I had a lot of mental illness problems and my mother applied for social security disability for me, so as someone wanting that just for mental illness, I had to take it as part of the adjudication process. The Feds at the SSA obviously don’t talk to the MRS people at all, because I took effectively the exact same test within a matter of months. It caused me to look slightly smarter on the second one because I already knew what they were going to ask for. There was a little difference in the specific things they asked, and I didn’t get anything right I hadn’t before (that I’m aware of), but I was certainly faster which I’m fairly sure is a component at least some of the time.

There are a lot of myths about IQ scores as well. They are based on a normal curve model and simply used to differentiate people along that Bell curve no matter what they measure. The most common standard deviation is 15 so that anyone that claims they have an IQ of 160 is already 4 standard deviations out (extremely rare; the flip side is an IQ of 40 and represents an almost completely nonresponsive person with extreme disabilities).

The irony is that people that really do have an IQ that high would instantly understand that but it doesn’t stop there. You can quickly go from a claim that you are the person that has the highest IQ ever to graduate from your high school, to the smartest person in your city to the most intelligent person that has ever lived in your state in just a few points at the high end of the curve. Some of the claims are pretty funny if you understand statistics at all.

When I was in 7th grade, I had a science teacher who told the class that there was a market for turtle milk, which was used to develop important drugs, among other things.

I think your high school psychology teacher must have had the same sense of humor. :wink:

The Department of Education for the United States has very little to do with the practical day-to-day education of students in the US. They certainly don’t keep a database of students in the US, along with their scores on any particular test, let alone IQ tests (which most students never actually have administered to them anyway).

It should be noted that the fit of a Gaussian (normal) distribution and a standard deviation of 15 are purely an assumption which is not borne out by looking out the extreme measurements. The reality is that it is essentially impossible to discriminate IQ at the low end of the spectrum due to functional illiteracy while ‘standard’ tests have no real validity above ~140. Actual cognitive abilities for functional adults likely follow a power law distribution, with the bulk of the population being centered around and below some mean value and some small part of the population being at an asymptotically distributed tail, just as it is with height, shoe size, and athletic ability. This still begs the question of what the ‘intelligence quotient’ actually measures; IQ is only somewhat correlated with academic performance and very poorly correlated with fiscal earnings or other measures of lifetime performance.


I’ve taken a real one in the last ten years, as part of a psychological assessment. And yeah, it’s the sort of thing you’ve probably done in school or seen online, but one on one and apparently not timed.

That’s not how intelligence works. Without actually having studied statistics formally, a very bright person likely doesn’t know what a standard deviation is, and may not understand why it would be “standard.”

It measures your ability to score on IQ tests.

Almost all children are given a test for giftedness. While not an actual IQ test it is trying to measure IQ.

When I was in about 7th grade, the whole class got some kind of test and there was a chart on the cover translating raw score to IQ. Granted it wasn’t the kind of formal assessment that a real IQ test apparently gives, it wasn’t something. I memorized part of the chart, so I know what it supposedly showed.

I know that MENSA gives some kind of test. I never took it, but I have looked at their samples so I know the sort of questions they use.

In the Army, the rumor going around was that the GT part of your ASVAB score was your IQ. I remember being somewhat proud because mine was the same as Richard Feynman’s.

Since then I’ve learned that’s probably entirely bullshit, and so is basically everybody’s understanding of IQ who isn’t a working researcher currently specializing in IQ.

Well, that’s cute, but despite the bold statements in the article, the actual metastudies that it links to do not generally demonstrate very strong (>0.5) correlation between IQ test measurements and wealth or performance. There is no question that having an above average IQ is somewhat correlated to intellectual performance but is not a strong linear correlation (e.g. having a higher IQ doesn’t translate into greater career or financial success than someone with a lower but still above average IQ). What IQ measures in a practical, functional sense is an ability to perform the particular set of skills involved in taking IQ tests; specifically, pattern recognition, logical and spatial reasoning, et cetera, while the skills to succeed in many vocations are as much social and empathic as they are logical or analytical. There are a lot of vocational and educational assessment tests (such as the ASVAB and college assessment tests like the SAT and ACT) which fairly correlate to IQ (unsurprising because they use much of the same methodology but in a less formal or guided context) but while they provide a reasonable correlation to academic ability (in terms of metrics such as cumulative GPA) they are surprisingly mediocre predictors on ultimate career performance.


You honestly think you know more about intelligence testing than Stuart Ritchie?

What specifically did he say in that article that’s incorrect?

That sounds like a huge waste, since there are plenty of career aptitude tests that do a better job for that purpose and which are easier to administer than an IQ test.

Yes, but there are also tests which use completely different methodologies than standard IQ tests, but which are still highly correlated with IQ. For instance, you can ask a child in a certain age range (pre-literate) to draw a picture of a person, and track the amount of detail included, and that correlates quite well with IQ tests taken when the child is older.

To even be able to make this statement requires assuming some measure of intelligence which is more descriptive than IQ. No psychologist would claim, for instance, that a person with an IQ of 140 is twice as smart as a person with an IQ of 70, because we don’t even know what “twice as smart” would even mean. Nor can we say that 90 to 100 is as big a difference as 120 to 130. But we would need to have a scale for which we can say things like that to be able to talk about what the distribution “really is”. As it is, IQ is exactly Gaussian, because it’s defined to be that way.

On another note: It occurs to me that the correlations between IQ tests and other tests are easily misinterpreted. For instance, suppose that a person takes the SAT, and gets a score three standard deviations above the average. That doesn’t mean that the most likely value of that person’s IQ is 145. Most likely, the person’s IQ is still below 145, because you have to consider the prior distribution, which says that almost all people have IQs below 145.

Why are those two at odds though? The power law distribution I could see, a vanishingly small % of people play a disproportionate role in the cognitive elite. I saw a study even verifying this among college professors, who themselves are likely among the top 2% of the population regarding cognition. Around 80% are below average for the group, a small 15% or so is around the average but the top 5% rapidly show massive amounts of achievements in their respective fields with the 99th percentile being significantly more accomplished than the 96th percentile.

But why is a power law distribution of cognitive skills at odds with a bell curve of IQ?