Hacker News Reader

Hacker News

Comment by Foobar8568 | Hacker News Reader

Parent Full thread

Foobar8568

Moreover, you don't even need to pass an IQ test to join MENSA US.

https://www.us.mensa.org/join/testscores/

View on HN

I struggle to see how a list of accepted IQ tests supports the idea that you don't need to pass an IQ test to join.

There are other tests that qualify you for membership, it appears.

I predict that in 20 years or fewer, they will accept algorithms interviews.

Or LeetCode

It is a list of different aptitude tests, e.g. LSATs, GREs and so on.

Aptitude tests are IQ tests in a socially acceptable disguise.

Not exactly. There may be some correlation, but there are certainly people who can get in via an aptitude test who would not be able to get in using an IQ test. The testing methods are completely different.

A person can improve an aptitude test score with some appropriate training, but a higher IQ person will tend to score proportionally higher given the same training/experience.

Yes, but the tests don't account for a similar amount of training, and so do not provide an adequate comparison. The whole point of IQ tests are to present problems which the test taker sould not have previously encountered, thus measuring their ability to quickly devise a solution, and indicating their intelligence on a standard distribution vs the population in general.

> There may be some correlation

There is so much correlation that the correlation of an aptitude test with an IQ test is basically equal to the correlation of one IQ test with another IQ test. (Around 0.8) By any definition other than "it can't be an IQ test unless that's part of its name", the aptitude tests are IQ tests. They do not differ in function.

That's simply not true. The form and function of the tests are fundamentally different. The biggest difference is that you can study for one, but studying for the other is considered cheating. You can see my other comment for further clarification.

You can respond to facts by saying "that's simply not true", but it's not going to be an effective way of convincing anyone else or of developing a coherent model of the world. The facts are that aptitude tests are psychometrically identical to IQ tests. They have the same characteristics and reveal the same information. Someone's score on an aptitude test as is predictive of their score on an IQ test as their score on an IQ test is.

What you claim is counter to the intended uses of the tests and their very definitions. One measures innate flgeneral intelligence, whereas the other measures skill or acquired knowledge in a specific area.

Then show me some sources to support your argument, because repeating false statements doesn't further your point either.

https://www.theclassroom.com/intelligence-tests-vs-aptitude-...

https://en.wikipedia.org/wiki/Intelligence_quotient

>The many different kinds of IQ tests include a wide variety of item content. Some test items are visual, while many are verbal. Test items vary from being based on abstract-reasoning problems to concentrating on arithmetic, vocabulary, or general knowledge.

The SAT tests abstract reasoning (read a story and answer questions), arithmetic, vocabulary... It's an IQ test.

Then why doesn't the article list the SAT in with the other dozen or so tests? Nor any other aptitude tests?

Not all standardized tests are IQ tests. Just as not all tests involving reasoning, arithmetic, and vocabulary are not IQ tests.

https://prepexpert.com/sat-to-iq/#

https://www.nytimes.com/roomfordebate/2011/12/04/why-should-...

Mensa considers that scores from after January 31, 1994, "No longer correlate with an IQ test."

I haven't taken an IQ test since... A billion years ago, so maybe it's the same, but having taken the LSAT, you can certainly learn them. My first LSAT I ever took was 78 percentile, whereas by the end I was scoring 95-98 with relative ease.

A normal IQ person can improve their IQ/LSAT/GRE test score with appropriate training, but a higher IQ person will tend to score proportionally higher given the same training.

Makes sense. Seems like IQ is just fluid intelligence which mediates the rate of acquisition of crystallized intelligence.

It's more the fact that these are not all IQ tests.

That is not a fact; they are all IQ tests.

You really need to Google what an IQ test is...

First of all, over 6 of the test listed do not give results in IQ points. So they're not nominally IQ tests.

Second, and more substantially, here's an explicit clarification for one of them (LSAT):

"The LSAT is not an IQ test. It does not measure intelligence the way IQ tests are designed to measure innate ability. A person who is very smart can receive a low LSAT score."

You can find similar declarations for the others (non-IQ) tests listed.

The LSAT is not an IQ test. It does not measure intelligence the way IQ tests are designed to measure innate ability. A person who is very smart can receive a low LSAT score.

What's interesting in the context of the post is that apparently I qualify for MENSA on the basis of several test yet have had not the slightest inkling to even figure that out, nor absolutely any interest in joining now that I know.

Maybe MENSA is a representative sample of smart people, but I have a lot of friends and colleagues I consider much smarter then me and none of us are in MENSA, so I'm skeptical.

What I find interesting on that page is that the more recent (past 2-3 decades) college preparatory tests (e.g. SAT) are not accepted as substitutes to IQ tests. Shows the trend of dumbing down the tests, I guess.

That is odd. If the tests were simply dumbed down, wouldn't Mensa just raise the required score? The implication is that these tests have become less IQ-correlated. I'd be interested to see a comparison in the type of questions from the 70s vs today.

Responding in a vacuum, but the other explanation could be that these tests fail to discriminate at the high-end. If everyone with a high score is squished together, raising the floor just amplifies relative noise

Probably less reliable and more overhead. You have to constantly evaluate what the new equivalent score and who knows, maybe even the max score isn't good enough anymore.

That's exactly what they did - they have 3 bands of years for the SAT. Two year bands (<1974, 74-94) are accepted with different scores, and the more recent tests (94+) are not accepted.

The major change in 1994 was the allowance of calculators during the math portion. I don't know why MENSA couldn't make a similar score adjustment here.

Well 5 years ago, they would accept all of them. Hell, they would even take a GMAT score.

https://web.archive.org/web/20161104152027/http://www.us.men...

Reply on news.ycombinator.com