Research vs. Comp Sci courses for stats

[mathmollusk]

I'm a junior math major looking to ultimately get a masters or phd in stats and trying to figure out how to spend this summer. I'm between either trying to get into an REU to get some research experience under my belt, or taking some Computer Science courses over the summer. (Unfortunately, my schedule constraints are such that I can't fit Computer Science classes into my schedule during the year, but I have heard they are important for stats). Which do you think would be a better use of my time? (If it matters, I will have an opportunity to do research through an undergrad math thesis the last semester of my senior year, but I guess I wouldn't really be able to report this on grad school applications unless I take a year off after undergrad).

Also, one more question: I've heard that if I get a terminal masters degree in stats first (and do a thesis to show research experience), and then apply to PhD programs, I can actually be at a disadvantage for admissions into the PhD programs since undergrads going right into PhD programs are judged less harshly and aren't expected to have as much research experience as people who already have masters degrees. Is this true, and if so, (leaving out the question of funding) is it to my advantage to apply to PhD programs right out of undergrad rather than masters programs?

Thanks!

Edited January 4, 2019 by insupliquitous

[MathMajor]

Do you have any experience with any programming languages (python, R, C++, Java, ect.)? If so, I would definitely go for research. If not, sometimes many of these REU's will teach you a programming language since you may be required to use a certain programming language in your project (assuming you are doing applied research). Thus, I think it is a good idea to go for REU's assuming that you do get in since they are very competitive. Taking CS courses can be your backup plan.

[data_scrub]

Hello! Senior undergrad majoring in Math + Stats and minoring in CS (so take my opinion with a grain of salt!) My CS background has been very valuable to me, but as far as coursework goes introductory classes in CS aren't super helpful if you already know some statistical programming (R, python, SAS, matlab, etc). Later courses such as Data Structures and Algorithms have helped me improve as a programmer and mathematician, but I do not value those courses over any of my research experiences or internships. I would recommend to go the research route this summmer, especially if you can apply to an REU! One thing I regret is not getting a letter for my phd application from my REU advisor.

Speaking of which, I'm applying to 8 different phd programs for next fall, and have been accepted to one so far! If you think you want to get a phd soon, I've been told there's no need to get a masters, but that it still can be a valuable experience. Those who do better in masters than they did in undergrad obviously have a more competitive phd application

[Bayesian1701]

I would go ahead and apply for the REUs.  I agree with the others that the CS courses aren’t that helpful unless you don’t have any experience and they are in languages you are likely to use. 

Now onto your second question. You will be judged a little harsher with a masters and the masters probably isn’t going to make you graduate with a PhD any faster.  The exception is you do extremely well in a top masters program and get great research experience in your first year but that’s pretty rare.   You don’t need any research experience so I wouldn’t worry about that.  I would apply to PhD if that’s what you think you want.  A lot of programs will consider you for a masters if you don’t get into the PhD program. You almost always have the option to drop out with a masters after a few years.

[bayessays]

Do research, absolutely no question.  This is where you will get letters that help you.  Students with a master's degree will not be at a disadvantage, unless you did worse in the MS than in your undergrad.