Passed Exam P (on the 28th)

[aum.tanay]

You are very bright or right PJA

[Zeigy]

That's some solid advice you cocky bastard. I'll keep that in mind.

[PJA]

''' either you lucked out something fierce or you are a grade a troll.

Okay, to be fair, I am pretty lucky with respect to my background being incredibly useful on this exam despite not having planned to prepare for it until recently. I hadn't taken any actual statistics courses in college, but understanding propagation of error from the first physics lab I had to take for my physics major gave me a lot of knowledge about the normal distribution. Having to do labs in modern physics lab where we measured radiation gave me knowledge about the Poisson distribution, and learning about entropy in thermodynamics required learning how to count permutations accurately in a variety of different problems. Taking a graph theory/combinatorics class in the math department also helped with that.

But since you seem so mad, maybe I'll provide some useful tips:
-Being a math/physics double major is incredibly helpful, and I'd recommend at least taking some form of discrete math/combinatorics/graph theory type course if you can. It's typically offered in the comp sci or math department, and as long as you're somewhat studious and enjoy logic you can probably take it regardless of your major
-Poisson distribution is just the power series for e^x with x plugged in, multiplied by e^-x to normalize it
-Make sure you draw a picture for every problem that involves a joint probability distribution function. It's very easy to mess up the limits of integration if you don't, but it's incredibly simple if you just draw the picture.
-Take as much calculus as possible (in addition to taking multivariate, I took 2 undergraduate analysis courses, differential geometry, measure theory and functional analysis, I don't expect that people in other majors could do anything like that, but I guarantee that if you show up to some study sessions for multivariate calc and try to help people after you've taken the course yourself that you'll become much better at it)
-As for exam taking, read the question, and answer it only if you immediately are 100% certain that you know how to do it. Otherwise, skip it and come back. I completed roughly 15-18 of the problems my first pass through the exam, then solved all but one of the problems (which I had to guess on) after reviewing through the rest of the problems. I had about 40 minutes left over at the end, and this is how all of my exams go (and how all of your exams can go if you don't waste time trying to work problems that will frustrate you at the beginning of the exam)
-Oh, and for the last tip, you'll probably freak out a little after you only do one problem then skip 5 in a row until you get to one you know how to do. Just try to deal with it, and when you get to the end and have solved half the problems in 1/4 of the exam time, you'll relax and feel confident again.

[trsspidey]

''' either you lucked out something fierce or you are a grade a troll.

haha it happens sometimes! I try not to get too upset when it's easy for someone. Congratulations!

[zenkei18]

First year math grad student.

Studied just from the free SOA sample questions.

IMO if you know how to read a normal distribution table, know how to do integrals, and know basic combinatorics you'll be fine.

Having studied some poker doesn't hurt either.

''' either you lucked out something fierce or you are a grade a troll.

[Zeigy]

Well done. That sounds like some good advice for me for September exams.

[PJA]

First year math grad student.

Studied just from the free SOA sample questions.

IMO if you know how to read a normal distribution table, know how to do integrals, and know basic combinatorics you'll be fine.

Having studied some poker doesn't hurt either.