Returning to grad school after 2 year hiatus
June 17, 2013 7:22 AM   Subscribe

So I'm beginning a statistics PhD program this fall and I'm concerned that my math skills have gotten rusty since I haven't done anything related for the past two years. I've been working as an actuary since I graduated college but I don't do that much math--mostly a lot of programming. Has anybody been in a similar situation to me? How was the adjustment for you? I'm considering retaking advanced calculus and linear algebra during my first year (probably next summer before I take 2nd year advanced courses) just to refresh myself again. I'm aware some people may think this is kind of pathetic but I'd rather be safe than sorry. Besides, it's only my first year. Is this frowned upon?

P.S. I'm doing stats grad school because
1) I want to become a stats professor or get more interesting roles in industry,
2) it's the best chance I have to explore interdisciplinary problems involving biology and finance, and
3) I don't enjoy being an actuary...
posted by molamola to Education (15 answers total) 3 users marked this as a favorite
 
Get the textbooks that those advanced calculus and linear algebra classes use and start working your way through them. Once you've knocked the rust off, you'll be able to tell whether you need the instruction again.
posted by Etrigan at 7:47 AM on June 17, 2013


Response by poster: Yep, started working through them. I'll definitely need the instruction again, lol.
posted by molamola at 7:53 AM on June 17, 2013


Have you looked at your local community colleges? Might be a cheaper and quicker way to do it, if you can at least knock through one of those classes this summer. Barring that, maybe get a tutor once or twice a week to help you work through the parts you're finding difficult.

But mostly, you should talk to your program. They'll likely have experience with people in your situation and have recommendations on how to or even whether you need to get back on track in those subjects. Alternately, they may spend some time in the first year refreshing you and/or teaching you the way they want you to learn it (the people in my MPP program who had the hardest time with the "basics" were ones who were fresh out of other programs and needed to be re-taught the MPP way of doing calc and stats and econ).
posted by Etrigan at 8:00 AM on June 17, 2013


Response by poster: MPP? What's that?
posted by molamola at 8:02 AM on June 17, 2013


(Taking to MeMail so as not to clutter up the thread.)
posted by Etrigan at 8:09 AM on June 17, 2013


This will depend on your program, but i would expect the summer after your first year to be too late. Your first year courses probably require enough linear algebra and calculus that if you get through them you won't get a lot out of an undergrad level summer course. I'd try and talk to some current students who came in after a break and get their take.
posted by matildatakesovertheworld at 8:29 AM on June 17, 2013 [1 favorite]


In my first year of (math) grad school, a fellow student who had come from industry retook a lot of undergrad classes his first year (linear algebra among them), and was behind us for a while, and we did worry about him, but he got a very nice post-doc at an selective liberal arts school when he did finish. It's definitely the right thing to do from a learning point of view, but you might take a reputation hit early on with your fellow students.
posted by Elementary Penguin at 8:40 AM on June 17, 2013


Um, what kind of statistics will you be doing? If you don't have a good grasp of analysis, you'll struggle in your first graduate course on probability. Is it the kind of statistics program where you won't be asked to do much math at all?
posted by Nomyte at 8:43 AM on June 17, 2013


I would look into doing an online course this summer. For example, MIT has an Open Courseware site, and you can take MIT classes online for free. You don't get college credit for them, but that seems fine for you. And since it's online you can go at your own pace.

Here are their statistics courses: http://ocw.mit.edu/courses/find-by-topic/
And here are the stats courses on coursera, which was developed at Stanford: https://www.coursera.org/category/stats

I would hesitate to retake a class you've already taken. Probably a couple weeks into it, it'll all come back to you and you'll be bored. And it'll slow down your progress toward your degree. It might be fine at first, but it could make you really tired of taking classes sooner than you have to be. Getting tired of taking classes is a major complaint among PhD students, so do all you can to hold on to your stamina.
posted by pompelmo at 8:45 AM on June 17, 2013 [4 favorites]


There's a chance that your grad school has a "math camp" just for this purpose. Math camps are a few weeks of intensive math refresher that takes place in the summer before first year classes begin. If the stats department doesn't know about one, there may be one offered by the Econ department of by one of the groups in the business school.

Also: you should do this before you start your first year. All of your summers during grad school should be focused on research (plus, you'll need the math refresher for your first year classes).
posted by eisenkr at 8:47 AM on June 17, 2013 [1 favorite]


Response by poster: I'm pretty good at analysis but I never learned multivar. My college was lame and skipped it and went straight into analysis. Also I learned abstract linear algebra but never how to apply it, so I was thinking about taking multivar and linear algebra (with matrices) at some point just to fill in the holes.
posted by molamola at 9:13 AM on June 17, 2013


Response by poster: Also is it normal to take three to four classes a semester?
posted by molamola at 9:18 AM on June 17, 2013


Yep, do linear and real with a focus on proof techniques before you show up. Additionally, some tricks from PDEs and complex are often useful.
posted by a robot made out of meat at 9:35 AM on June 17, 2013


Well, PhD programs vary. I would try to get in touch with someone at your school - definitely your advisor, and perhaps a 2nd or 3rd year graduate student as well - to ask some of these questions. Things like "how many classes should I take each semester" can really only be answered by people at your school.

One thing that might help is reminding yourself that you got in to this program. They want you there, and they want you to succeed. Talk to your advisor, and see what resources are available for students like you. You're not the first person in this program to have this question, I'll bet.
posted by k8lin at 9:44 AM on June 17, 2013


Stewart's Calculus textbook is very good. You can review the differential and integral calculus rapidly and then the multivariable section will follow quite logically. I'd also recommend Ross's "A First Course in Probability" for review.
posted by pjenks at 1:36 PM on June 17, 2013


« Older Is there an application that supports my...   |   What awesome contraband can I give my newly-not... Newer »
This thread is closed to new comments.