Applied Stats: How much math do you need?

[cooperstreet]

Hi All,

 

I'm an incoming PhD student in political science that hopes to do a lot of quantitative/applied statistical work. Previously, I was studying to be an actuary and passed two exams, so I have a firm background in doing calculus problems by hand.

 

My question is, beyond calculus I, II, and III, and linear algebra, how much more math will I need to know to do sophisticated applied statistical work? I know that I will have to understand the underlying concepts, but will I need to know how to calculate a lot of these equations, especially since I can do them in R? Thanks.

[bsharpe269]

You should be fine in my opinion... I recommend taking a great advanced statistics course in grad school if you can find one. To understand the theory, you obviously need a strong background in proof based math courses. I was a math major as an undergrad though and many of the professors who I see usuing lots of statistics (and understanding what they are doing at a conceptional level, not necesarily at a rigorous mathematical level) have less background in math than you.

[cooperstreet]

Thanks for your help. Can you give me an idea of what you mean, re: type of course? I may audit it rather than formally enroll. I'll have a lot of statistical training in political science, but its obviously looking at applied research.

[reinhard]

I think to truely understand statistics and probability theory, you must take measure theory which requires real analysis.  But some may disagree with me because in my department, there are many MSc stats students who don't know a thing about measure theory or real analysis.