alt Hacker NewsBefore anything one should probably check or at least ballpark their IQ score. The median IQ for mathematics PhD students probably hovers somewhere around 145, about the top 0.2% of the population, correlated with about a 1510/1600 on the SATs, a 34 on the ACTs, etc. Those aren't perfect correlates but you're much more likely to have an SAT or ACT score than a professional IQ score handy.
Math is infamously g-loaded, pure math even more so. An unfortunate fact of life. On the bright side, math is very much a "shoot for the moon and you'll land among the stars" subject to pursue if you even loosely keep industrial or business applications in mind.
Replies
Two things :
> The median IQ for mathematics PhD students probably hovers somewhere around 145
Does that mean the 145 figure is only a guess on your end ?
Second, as far as I know, an individual's IQ is not something set in stone, and can absolutely be improved with training. I remember reading (that's an anecdote so correct me if I'm wrong) that rewarding a good score with money was able to improve the outcome by up to 20 points. It doesn't sound absurd to me that someone with a slightly above average IQ could get close to 140 after 6, 7 years of high level math training.
My experience with pure math is that this is not necessary to get job as one, even at good institutions, but you will be terrorized by the arrogance of the ones you mention. Learning to deal with the "brilliant jerk" is a problem in many fields, but the ones I've met in pure math are some of the weirdest (and most vicious)
What's much more sensible than taking an IQ test is looking at your experience with math to date.
Or, alternatively, you could just skip over "general intelligence" and get to your second measure that's directly validated against college outcomes – SAT score.
(My understanding is that the "general" GRE – not the math subject test – is less of a predictor of completing a PhD, but I think we can come up with a hundred reasons why.)
[1] https://journals.plos.org/plosone/article?id=10.1371/journal...
!!!! IMPORTANT COMMENT !!!!,
because the edit button is now gone:
I misspoke, 145 should have been my loose estimate for the median of actual working mathematicians and taken with many more grains of salt. Mathematics PhD students cluster around a much more attainable average of 128-130. Per [1] this would map to a much easier SAT score of only around 1280.
My general point still stands that you probably want to look at this and consider your potential career in math against this, but the skill curve is less punishing than I initially thought.
But why would you do an IQ test, when you could just do a math test? Surely a math test is a better indicator of math ability than a test that is merely correlated?
The interest in pure mathematics probably sets inquiring individuals in a higher intelligence bracket already. If somebody got through high school enjoying and succeeding at geometry and calculus, then they probably could stomach most undergraduate work in the same manner.
It seems a bit like gatekeeping to make people question whether they are smart enough when they will figure out pretty quickly if they have the aptitude or will to do it just by being exposed.
Imagine if Richard Feynman used his IQ as a metric for deciding whether he should become a physicist. Physics would not be the same.
I am certain that there are mathematicians below, near, and above an IQ of 145 that all have great research productivity. IQ tests do not approximate the creativity, effort, and collaboration required in a mathematician. Not to mention the dubious nature of the 145 claim.
Of course, there are some people that will have a greater aptitude for mathematics than others. But you do not need to be a genius, and this is echoed by Terence Tao [0].
[0] https://terrytao.wordpress.com/career-advice/does-one-have-t...
> "shoot for the moon and you'll land among the stars"
I´ve never been able to wrap my mind around this saying.
Unfortunately, I have heard logic like this throughout my life, leading me to decide that because I struggled with (some subset of) math, I am not an intelligent person, which led to forcing myself to do pure math in college to prove my intelligence. This led to many significant and awful mental health issues. While this is a bit of a fallacious logical leap, it's not impossible that other people have went through this because of this sort of information being hammered into their head.
I choose to believe succeeding at anything is mostly about persistence and interest, barring other immense structural factors. I have zero interest after doing difficult pure math classes, so I stopped. I now think I am good at what I do, but everyone's intelligence and interests are different.
I think this sort of quantification of intelligence is really harmful to people. I don't want to exclude people from pursuing their interest because their SAT score wasn't high enough. I have met math PhD students with bad GPAs and poor math class grades in their undergrad.
On a tangent, CS undergraduate programs are insanely competitive and filter in crazy ways, and most of my friends who were passionate about CS (especially systems CS and SWE) did ECE just to avoid the competition and dispassionate culture. Your GPA and SAT scores had to be insane to get into almost any undergrad school for CS.
I was about to ask if there is any way we can bridge the mathematics the same way we did with programming ; I love going-on with an LLM to learn math and apply them directly to problems I'm facing ; and to be honest I often feel like a musician who never actually learned his tuning lately ; especially in topics correlated to balancing, monitoring, and even simply in projection of cost ; It's like I have been over-focused on complexity of algorithms without actually realizing that it's only one part of the problem - so I have a huge potential usage of mathematics and one can really easily leverage them thanks to AI - especially considering tests ; but that's where my road ends without a more sophisticated approach - and even then it's a very dark place to wonder by (eg: lot of time spent for seemingly unknown appliance) it does however start to feel even more attracting for these exact reasons - I feel like having problems to solve with just makes this a lot easier - but I'd love to be able to ground-up things even more - and especially be able to take shortcuts.
I mean a lot of people just run a database but don't know wtf it does - but it still useful to them - maths however need to be understood to be really useful -
Is there not a way to make this lot more navigable ? Are there bridge concepts that are important enough that we can spend some time to learn them ? (there are ofc) - and how deep shall we go ?
I did consider doing a mathematics postgrad qualification but my IQ is high enough to realise I liked actually getting paid decent money.
Not true at all, we need mathematicians and scientists from all backgrounds, and creativity comes in many shapes
I got rage baited by this so hard, cant comprehend thinking this way.
Hung out with PhD's, economists, bankers, trust find kids, scientists, and artists - who maybe weren't top tier enough, but none thought this way.
Literally the weirdest take on a forum filled with dreamers, but every take is valid.