University Selection for MS in Statistics/Data Science

The application season is upon us and Santa’s about to decide who’s been diligent with their applications and who’s just been plain lazy.

I’m in my final semester of MS Statistics at the University of Illinois at Urbana-Champaign. Like last year this time, my Quora and LinkedIn inboxes are full of:

1. Hello ma’am, what to do to score good marks in GRE?
2. Can we prepare for GRE in 2 months (1 month/ 30 days/ 15 days/ 5 days…what?)?
3. How did you choose the universities to apply to?
4. How’s the UIUC MS Statistics course? How is it in terms of giving students an exposure to data science?
5. How are the Data Science courses different from Statistics/CS courses?
6. How is the MS DS course in Columbia/Rochester/Minnesota?

I wondered about some of these questions about 2–2.5 years ago when I was applying, so I understand how confusing and daunting the whole process can be. And I’ll try to make the procedure clear in this blog. BUT that being said, I need to state this. Whenever you plan to ask someone questions about their application process, here are some things NOT to ask:

1. Can you email me your SOP?
   - No. Write your own damn essay. It’s an autobiography, not a biopic. Don’t go looking for “inspiration”. An “inspired essay” is a copied essay. That easy. Look for some sample essays online and get started if you need to.
2. Can you tell me about your experience while applying?
   - Could you BE more vague? Be VERY specific about what you want to ask. Some of us are still studying and are battling close to 2 assignments, 3 quizzes and 1 exam every week.
3. Which universities did you apply to? How’s the course at *insert name of university that I did not apply to*?
   - How would I know? If I’d have deemed it good, I’d have applied for it. If not, it was either not in my preliminary list (we’ll get to this later) or it was and got filtered.

Now that that’s out of the way, let’s try to streamline the process.

NB: This whole blog is based on my personal experience and may differ from person to person.

GRE and TOEFL

I wrote an answer to Quora question (which, by the way, is a great resource if you want to learn about other people’s “experiences”. See Q2 above.) about my GRE prep — Quora answer. I basically had 1 month to prepare and did that while I was working. This answer also goes through some tips to help you stay motivated. I scored 330 ( V: 163 Q: 167 AWA: 5.5). But there are people who score 340 and I would strongly encourage you follow their prep too. Be as high while aiming as possible.

Since you’re applying for Data Science or Statistics, I’m guessing your math is already strong enough for GRE, hence try to focus more on Verbal. Also try to give your TOEFL in the vicinity of your GRE exam. Helps because your verbal prep will be carried over. My TOEFL exam was 4 days after my GRE and helped me stay on track.

For TOEFL, all I did was watch Notefull videos and complete the mock exams. That’s it. I scored 115 (Speaking: 28, Listening: 28, Writing: 30, Reading: 29) and that was just from very little prep (because I was lazy and was already very late in giving this exam; I gave my TOEFL on December 17, which as most of you would be aware, is already past some of the deadlines)

Making your preliminary list

To choose your universities, you need to have a preliminary list of universities that shall be your lookup. This list has the following attributes:

1. Contains all the universities that you’ve heard of that are offering a DS/ Statistics program or universities that you are considering to apply to.
2. Has details like the rank of the department (for Data Science courses I recommend having rankings of Math, Statistics and CS rankings as usually the DS curriculum is just an amalgamation of courses from these departments), the fee and the cost of attendance (which includes your living and food expenses), location et al. Basically all the factors you would like to take into consideration.
3. Is exhaustive i.e. once you’ve made this list you DO NOT go changing it or choose a university that is not in this list. It’s an immutable, exhaustive set of universities. This is important for you to be organized, which will be a benefit later on when you choose your final list.

Looks something like this:

My prelim list was something like this. You can use this for reference.

How to make this preliminary list?

1. Visit the websites of each university’s department. Research through their course, other related departments, research facilities available, scholarship (tuition waiver) options, faculty etc. This is the MOST IMPORTANT step. Ideally this should take about 10–20 days of time if done properly. Imagine reading through and critically reviewing 10 research papers.
2. Read blogs and check rankings here and here.
3. Ask alumni and current students of those universities any questions you still might have. But remember — be specific. A good place is LinkedIn, it’s easy to search for alumni and students there. Facebook and Tinder are a strict no.

Making your final list

Choosing your final list of universities to apply to depends on your financial resources and your profile. The groups on FB like this Yocket group or this one can be a useful resource to see where people of various profiles are applying. A good option is to search for some keyword like your GRE score or CGPA and try to find the closest profile to yours and then seeing their post and comments. Be creative. Try not to DM the person whose post you’re looking at. I’m sure they would have already gotten many messages. Just leave a comment for them to get back to you.

Yocket is an app that lets you see this as well. You can also compute your chances of getting into a university. Though I did not use these features, I did use Yocket to send my transcripts to universities. Cheap, streamlined and reliable. I think it used FedEx as an external service, but I’m not sure what their procedure is now.

My final list was based on these features (in no particular priority):

1. Combined rankings of Math, Stats and CS departments (I used a weighted ranking system, where Stats > CS > Math )
2. Cost of attendance
3. Flexibility of the course and the curriculum
4. My irrational desire to apply to highly ambitious colleges

My final list (without results) was this:

Spreadsheet can be accessed here

MS Statistics at UIUC

When I received all my results, I was confused between two universities — Columbia and UIUC, both for MS Statistics. I made my decision solely on the fact that the probability of getting financial aid in UIUC was higher because my seniors had received an assistantship (which is basically your tuition waiver and pretty much all you need) whereas in Columbia you would get a grader job or a TA job but that was paid hourly and hence would amount to about $2k per semester (with what information I got from the students). Given the difference in standard of living in NY vs Champaign, I chose UIUC.

And I’ve not regretted my decision since. Here’s why:

1. Flexibility: There are 2–3 core courses (See this), but the rest of them (required credit hours are 36) are elective. There’s a long list of courses that you can choose from and even swap some electives for some CS courses. I’ve taken NLP and Deep Learning from the CS/ Industrial Engineering departments. I’ve even audited a Data Structures undergraduate course because it’s highly popular at UIUC and I felt like I needed to relearn the basics (has about 800 students every semester).
2. Financial aid: Apart from the first semester, I’ve held a 50% assistantship appointment every semester, which means my tuition is waived and I get a monthly stipend (depends on the department you work for). This means a very high ROI for the course. Assistantships are offered by other departments (not Statistics) and so you have to fight for them. But from what I’ve seen, they’re not THAT hard to get for us since you basically see at least one Statistics course in all Science and Engineering departments.
3. Internship: If you choose MS Statistics Analytics , you have a mandatory internship to complete the degree. However you can also do it in MS Statistics (or do Statistical Consulting which is basically giving statistical advice to companies looking for data science/modeling solutions). Our career fairs are huge and offer a lot of opportunities for students to find internships, co-ops and jobs. We usually go to Engineering, Statistics & Math and Business Career fairs. That’s 3 events per semester and that means a whole lot of networking. I got my internship offer by applying through our university’s online portal.

This was my experience. Hopefully it helps with your procedure and selection of universities. Again, these are my own experiences and stories. And I wasn’t sponsored by any of the resources I mentioned (though I’d like to so Yocket people, if you’re reading this..!)

And lastly, while university selection may not seem important while giving your GRE exam or looking at the whole procedure in general, in my opinion it’s the MOST vital step. And it may seem daunting at first since it’s very easy to get lost in such an ocean of information (you might use BFS for visiting the university website, but trust me, use DFS — gives you the required info faster), but being organized and making Excel spreadsheets is the key (as my father used to say). Make notes, write them, type them, upto you. You’ve got this.

Let’s make America great again. (Trump, if you’re reading this, I’m on your side. JK. Study in the US but as my actual sponsor Modi says, Make in India.)