Thanks for your answer,
" Regardless of whether you fall on that spectrum, I think it's a terrible idea to actually use your IQ to decide whether you should study math- no matter how good a predictor it is, your actual math performance is an even better one. "
Got to agree on that first part, I thought the same when I was younger ; I've actually already started studying. My biggest issue (and the reason behind this) is that my work and investment (3rd year undergrad math, at a very good french university) are seemingly unrelated to my results.
I've went to doing almost nothing and somehow managing to pass, to trying my best and having apparently worked my ass off and failing. Results which have been average,mediocre,good.
If I'm looking for another (perhaps not so related) predictor of math performance it's because I feel like at no point any work I did actually paid off in proportion to my efforts or even had a positive effect on my results.
Proofs aren't the problem, I tend to forget things and am rather intuitive in my functioning so I usually lack in rigor/tend to miss details while writing.
I've invested enough not to bother anymore (and I'm pretty close to letting go, even though I love maths) if I don't have at least one good reason to keep going (I've already went much further with compromises than I thought I ever would).
I will probably go and do the tests, if the results are high enough I will most likely give it one last shot, otherwise I will be done with it. (probably more about having an objective reason than what the tests themselves mean)
As you said " no matter how good a predictor it is, your actual math performance is an even better one ", but deep down I'm still not convinced, how irrational it might be.
I meant the "Perceptual reasoning index scale" in particular ; I don't really care about verbal score as I doubt it could be a limiting factor for me. The 2 others indexes (working memory, processing speed) might be important but I will give them importance only if the score for perceptual reasoning is already high enough.
(using the WAIS-IV, performance IQ is indeed an older concept)