Raven’s matrices | Hacker News Reader

Raven’s matrices

sep.stanford.edu

38 pointsandsoitis32 comments

> Seriously though, what should be our relationship with those fated to be born less gifted than ourselves? Two ways to look at this: (1) What should USA race policies be? (2) In your declining years you write a will to disburse your estate. Assuming your children vary widely in intellect, health, wealth, and honesty (as countries do), what should your policies be?

It's wild to me that people view the world this way.

yikes

Yeah what's up with that? I was reading the page and it seemed fun, then suddenly it goes into some rant about, I think, immigration? Not sure what a "race policy" is even supposed to be.

It says > A newer test (complete, free) is at https://iqpro.org/. This site asks you 60 test questions, then asks for your personal data, then charges 20 Euros to show how you scored.

The justification on the main page doesn't inspire confidence:

> IQ test reports costs 19.99 due to increased costs associated with collection, updation and normalization of data.

The real IQ test is whether or not you'll pay $20 on that justification

What if you are very reach and $20 means nothing to your (I myself poor, $20 is big deal for me)

Take this test instead: https://openpsychometrics.org/tests/FSIQ/ It's free and IMO the best of the free tests.

I'm skeptical, given that first section is testing vocab -- not "infer what this word means from context", but just straight-up "how many words do you know", which I don't think has much to do with intelligence (whatever that means).

At the end you'll get separate results for Memory IQ, Verbal IQ and Spatial IQ in addition to the overall score. So if you don't think knowing words has to do with intelligence, you can just ignore the verbal part.

I'm skeptical of other parts of it as well; for example, one of the memory sections uses logos from popular software projects (inkscape, gnome, etc). It was a lot easier to remember the logos I'd seen before vs the ones I hadn't.

I think it's a neat test, and I'm curious to compare results with my friend who's an artist (I found the visual components challenging). But it seems less robust than something like ravens matrices.

135 can I get free money or a job now

It is a gullibility test masquerading as an IQ test.

I found a link to the original standard question set:

https://pdfhost.io/v/iaIChY.6O_Ravens_Standard_Progressive_M...

There are two other sets -- a colored set designed for children and an advanced set for older teenagers and adults.

Thanks for the link. I like taking these for fun, any idea where to find the advanced set for older teenagers and adults?

I found a couple copies:

https://drive.google.com/file/d/1QlyZkyy8wKkcVcFNB8pf1uslgEu...

https://web.archive.org/web/20170707220206/http://splushka.c...

Source: https://puzzleaday.wordpress.com/2022/09/25/the-ravens-advan...

The author has a wonderful book (https://sepwww.stanford.edu/sep/prof/fgdp/toc_html/) that's a mix of physics, signal processing, statistics, and matrix algebra.

I wonder, why he has refused to do the last test it looks pretty easy... It is an addition test with loops directed inside are negative numbers and loops directed outside the positive numbers. Or I'm totally wrong and something different is expected ...

I don't know how IQ tests are supposed to be conducted, but when I first heard of Raven's matrices in middle school, I jokingly asked all my friends to do it. What struck me at that time was that for almost all of my friends, there were no questions that were too difficult to be solved, but there was a pretty large variability in the time taken to complete the test. For some of the friends, answering the questions required a lengthy logical deduction process; for others they almost picked the answer just by intuition and then justified it afterwards.

At that time my takeaway was that an IQ test with a generous time limit probably meant nothing, but an IQ test with a strict time limit could probably test intuition.

That said the utility of testing intuition still isn't clear to me.

In order to make complex decisions in novel circumstances involving many factors under time pressure, as many businesses do, said intuition helps exhaust terrible solutions, and see potential solutions.

I found the last question quite easy.

I have a set of books of these and some of them are much harder.

There are also the similar bongard problems.

Yes, the whole test seemed quite easy, but then in the end I only scored 129. I guess if I spent more time on it, and haven't been mildly drunk I could do better.

Also, "raw test score" was 57/60 (so I got only 3 questions wrong). Given that, I wonder how they scale scores from 129 to 140 and beyond - do they only count correct answers, or time taken as well?

Also: it's a pity they only support PayPal as payment method - I needed to create throwaway account just for this one transaction.

So is that last one just a simple +/-, or (like a lot of these tests) there are alternative solutions and I've simply got the "wrong" one?

Flynn effect. It's indeed trivial.

> This is the last question in the test. I did not even try it. Enjoy

Maybe 2? I can only guess what the answer is. I don't the rule of sequence here?

Just add the polarity of the loop counts (my solution, no idea if it's the right one).

I got 5: 1 inside + 3 outside = 2 outside. The other two rows look to be 4 inside + 1 outside = 3 inside and 3 outside + 2 outside = 5 outside.

Just addition with inside positive and outside negative (or vice-versa)

It is 5 [edit for clarification: the answer is choice number 5, the numerical answer is 2]. Each image represents a positive or negative number. Adding the first two items in a row gives you the third. Alternately, adding the last two items in a column gives you the first.

so it's 5, not 2. an outer and inner cancel each other out. then the remaining whether inside or outside are added to that same side.

Sorry, I was mixing up the numerical answer and the choice number. The answer is choice #5 in the list; the numerical answer is 2.

It seems like too many people score well on this test for it to be useful anymore , due to the Flynn Effect or other factors. I got a top score

Reply on news.ycombinator.com