Carbon Offsets!

For each view, 50g of carbon is offset by the Bonneville Environmental Foundation.

As of 2019-09-09, that's 3,134 kg / 6,903 lbs!

Wednesday, 10 October 2018

Applying for a Master’s Degree in Stats or Data Science – Why and How

Why apply for an MSc, instead of finishing with a BSc in Stats or Data Science?


Statistics is a ‘discovered major’. Traditionally, not many people have gone into university planning to be a statistics major early-on. It also relies, as an unofficial prerequisite, on a wide range of skills that are typically learned in an undergrad degree.

For example, a statistician is expected to have a background in mathematics, in writing and communication, and in programming. They’re also expected to know a little about their respective service or collaboration fields.




For example, a health sciences statistician should know something about health and biology to understand the data that they are working with; a statistician working at a large company should know something about finance or marketing or customer relations if they are to be able to explain and use their findings effectively. Realistically, this is just too much to learn in a four-year major, and as such, many statistics students opt for graduate school.

Applying as a BSc Statistics student


- Luckily, stats students typically have a very high success rate for graduate school compared to undergrads from other majors. It’s by no means a guarantee, but it’s better than the 10% (or lower!) application success rates that some other majors suffer.

- A lot of statistics grad school applicants come from fields outside of pure stats too, so having a stats undergrad degree gives you an advantage from the start.

- Furthermore, grades matter, but they are far from the ONLY thing determining your success. Once you have the required minimum (usually 3.50 GPA or 3.67 GPA, occasionally more), other factors matter much more. (Here I refer to an A as 4, a B as 3, and so on.)

What if my grades aren't that good, or I'm only a minor in stats, or my English isn't good?

This is a setback, but it's not the end of your story.

Think of yourself as an RPG character and the MSc application as a big boss. If you can't succeed now, it doesn't mean that you can't succeed forever. "Don't be hasty, raise your levels." (Zelda II: Adventures of Link manual, 1986)

If you apply to some degree programs, but don't get accepted into either one, or can't go this particular year, that's fine.

- Go make some money,
- get some life and career experience,
- compete in Kaggle or Topcoder competitions,
- participate in local hackathons,
- get a MOOC (or SAS programmer) certificate or two,
- take some additional undergraduate math, stats, or computer science courses as a continuing / legacy / mature student
- build a portfolio on GitHub or one of it's competitors,
- start a blog,

...and then try again in 6-12 months.  If you still fail to get in, you've at least built up your earning power, which is what an MSc would do for you anyway.

There's a myth that any delay or lapse between graduating a BSc and the planned start of a MSc is looked down upon, or even that taking more than 4 years to get a degree is going to kill your chances at going further, and that's just not true. Much closer to the truth is that all that time should be accounted for - if you don't have the funds or family support to study full time, that's accounting. If you take six years because you changed majors, you have the extra credits to show for that. If you take one or two or five years between degrees, be prepared to write about your growth and learning and acquired skills that happened in the meantime.

I took five years to finish my BSc in Math because I moved cities and schools, and couldn't transfer everything. After that I took a year off to work as a tutor and try (and fail) to start a career in finance. My GPA was about 3.4. Then I applied to do an MSc in statistics, and in computer science and was tentatively accepted to both.

Don't be left out in the cold, build your data skills!

There are two types of Master’s Degree programs: The coursework MSc and the thesis MSc.


- A MSc typically involves 12 months of intensive graduate level courses (three semesters, including summer), and 3-4 months to work on a capstone. A capstone is like a well-researched term-paper. A coursework master is the faster, more efficient option for those looking to learn advanced degree-level skills and get into the workforce quickly.

- A thesis MSc involves 12-16 months of grad level courses and 8-12 months to write a thesis. A thesis is an extended research paper. Compared to a paper that you might find in a research journal, a thesis includes more background information and programming details than you would normally have. A typical statistics thesis is about 60-80 pages long (there’s more work behind the scenes that doesn’t show in the actual writing, compared to, say, an English MSc thesis, which can run over 200 pages). A thesis master’s is the deeper, more comprehensive option for those looking to spend the time to get all the skills necessary for both industry work and PhD level research afterwards.

- For most thesis MSc degrees and some coursework degrees, you are applying to both the school and to a particular supervisor. Every year (or intake period), the applications are pooled together and summarized, and faculty that are looking to take on new graduate students go through the application pool and pick ones that are promising to offer to supervise. (At least, that’s how it works at SFU)

Maximizing the chance of success.


- The culture and systems in place at every school are a little different. For SFU and UVic, it’s best to contact the department and your preferred supervisor faculty. For UBC, they just want your application, and specifically do not want you to contact their faculty. At SFU, you get assigned a supervisor before you start, while at UBC, you get one after your first year of graduate studies.

- A lot is balanced on the cover letter you’re allowed to send it. Get it proofread by multiple people, including someone already in the field, ideally.

- Just as if you were preparing to go straight into industry, extracurriculars matter a lot. If you have experience at a co-op, that helps. If you have experience doing research through something like a USRA (see NSERC, the Natural Sciences and Engineering Research Council) or a similar undergraduate research project (see SURJ, the SFU Undergrad Research Journal), that helps a LOT.

- Consider your references carefully. The best references are the ones that personally know you through exposure to you as a research student, or in a SMALL, upper division class room. For example, when applying to stats grad school, your Stat 410 prof would be able to provide a better letter for you than your Stat 101 or 270 profs or your History 404 prof. Likewise, professors will be able to give better letters than graduate instructors or new lecturers.

- References are often asked to fill in forms about you about your ability to complete research, undertake advanced studies, and COMMUNICATE both in writing and verbally. If you’re struggling to communicate verbally in the official language of the university (overwhelmingly English or French in Canada.), then find a conversation club through a site like Meetup, or join a public speaking group like Toastmasters.


Some other resources:
Comparing costs of living for choosing a city:
https://www.numbeo.com/cost-of-living/comparison.jsp

UBC's Tips on getting scholarships:
https://www.grad.ubc.ca/scholarships-awards-funding/resources-award-applicants/tips-best-practices

From:
Washington Post’s article on grad school application success: https://www.washingtonpost.com/express/wp/2015/02/09/outshining-other-grad-school-applicants-is-as-easy-as-1-2-3-and-4-5-6/

Top Universities application tips, by Laura Bridgestock:
https://www.topuniversities.com/student-info/admissions-advice/grad-school-application-tips

No comments:

Post a Comment