What I wish I knew before my Biostatistics MS

Graduate school and college are not the same thing

Book summaries for learning faster (Sponsor)

I read a lot. But I need to be picky with what I read. I want to make sure a book is worth it before I commit my limited time to it.

Thanks to Shortform’s book summaries, I can quickly grasp the essence of a book and make that judgment faster. Shortform creates high-quality summaries that cover the key ideas 1000+ books. Even if a book doesn’t meet the cut for a full read, you can still learn from the wisdom catered by Shortform.

If you join through my link shortform.com/verynormal, you can check out Shortform with a free trial. If you like your experience, you’ll also get an additional 20% off an annual subscription through my link. Give it a try, will ya?

---

In this issue…

I’ll talk about a recent chat I had with an incoming Biostatistics MS student. It made me think about how I struggled as an MS student and how I eventually adapted to the difficulties of the program. As I approach the last year of my Ph.D, it inspired me to document this advice while it’s fresh in my head.

Here are 5 things that I wish I knew before I started my degree:

1. Grade inflation makes standing out harder

Having been in two different universities, I’m comfortable in saying that grade inflation is definitely more prevalent as a graduate student. If you don’t know what it is, it’s the phenomena that more of the grades will be high B’s or low A’s. There are a lot of reasons for this, and they’re not uniform across school or faculty, but here’s a few things I noticed:

• Some students will just complain about the grading a lot

• The professors and TA’s don’t have the time to deal with angry students

• It looks bad for the professor and the university if too many people get bad grades. There’s not much time to retake classes, so might as well pass them

The grades themselves don’t matter so much unless you’re looking to get into a Ph.D program right after you get your MS. But even then, the grade inflation makes it much harder to stand out against other MS applicants. To get a 3.5+ GPA in graduate school is not that impressive, it’s almost the norm.

That being said, this might be different for other schools and degrees, so always take this knowledge with a grain of salt.

From my personal experience, I thought it helped demonstrate that my attitude towards school had changed. I was a low B student in college, but I demonstrated through the change in my grades that I was serious about being a Ph.D student.

2. Stand out by (independently) seeking work

Since grades aren’t the proper way to stand out in a graduate program, then what is? How does one acquire those crucial letters of recommendation or get access to interesting research projects?

The answer is to put in the work to cold email professors and ask them directly what they do. At Columbia, there were ~100 MS students and a biostatistics faculty of about 10(?). To stand out among the crowd, you have to actually read into recent papers that the faculty have put out and demonstrate that you put in this work. This will take longer to vet projects and professors, but you are much more likely to start a great working relationship if you invest that time.

I understand that some students may be afraid to talk to professors directly. They end up writing formal emails that sound stiff and robotic, and it’s hard to tell how you’d work out as an analyst/researcher. But I have to emphasize that a MS is only 2 years long.

That is not a lot of time to get a working relationship started and put out meaningful work. You have to put yourself out there.

Sometimes, the administration may send out emails about specific opportunities in the department. While these are nice, I would advise that you don’t wait for these to come out. Literally everyone in your graduating class is seeing the same email, so competition will be high.

3. Wait to learn your fundamentals before seeking projects

I got this tip from a talk I had with my would-be boss in the department:

Wait a year before you start asking to work with professors. Before then, you’re just not equipped to be helpful.

I know that some ambitious students will want to hit the ground running when they get into the program. I was one of them, and I didn’t listen to my boss’ advice in my first quarter.

I found a random position with a researcher in the medical school and she gave me a project to do. Little did I know was that she was expecting me to understand all the statistical details because she herself was new to the methods.

It goes without saying that I had no idea what I was doing, and when this researcher realized it too, she chewed me out in the middle of the study lounge.

This advice especially holds if your undergraduate background is not math or statistics. If so, graduate level statistics will be a large step up from what you are used to in terms of math rigor. I was an engineering major, so I was familiar with the use of calculus and linear algebra, but it was still very difficult for me.

I know in the last point that I mentioned you should be contacting professors you want to work with on your time. Points 2 and 3 may seem in conflict, but let me clarify. While you are learning the fundamental concepts and models in your first year of classes, do your research into who you want to work with. Read the papers, but don’t force yourself to read all of it; just get a gist through the abstracts. When you actually meet with the professor in person, you may actually find that they are working on totally different projects, so it’s best to be prepared to handle what they may have.

Caveat: if you have a very strong math background already, I can see someone starting early. But, among my peers who did come in with a background, not many were much more successful with projects compared to those who waited.

4. Strive for deep intuition, not a good grade

A biostatistics MS can be a terminal degree. This means that after your MS, you are theoretically equipped to start working in statistician positions in industry (and data scientists, etc.). You are expected to have a solid understanding of the statistical ideas and be able to apply the most common tests without much supervision. Of course, there will most likely be a more senior statistician to help, but you can’t totally rely on them.

I’ve seen that some of my peers in the program treated the MS as an extension of college. They were just getting the homework done, cramming for midterm and final, and letting all that knowledge spill out of their head afterwards. These people were having a much harder time finding jobs after graduation.

Others understood that a technically-minded job was going to come after their MS. These people were common visitors to the office hours and were not afraid to clarify on points of confusion during lecture. They sought a deeper understanding of the material, and sometimes this comes at a cost to grades. They may not have enough time to fully understand the subtleties of the Cramer-Rao Bound or Maximum Likelihood Estimation by the midterm, but by the final, they were ready for any curveball question.

These people had jobs secured before graduation or soon after.

Your employer will not care that you got an A vs an A- in linear modeling. They will care that you know when to apply it and when not to. Seek this latter goal, not the former.

5. Publications are not a dealbraker for applying to a Ph.D

This one is dedicated to those small few who are interested in applying to a Ph.D program.

There is a popular conception that you need publications to get into a Ph.D program. While they certainly help, they are not a dealbreaker. The reality is that papers take a long time to develop, and many of the factors that go into writing one are out of your control.

Admissions know this. What they want to hear from you is that you have:

• A proven record that you can handle graduate level math

• A solid reason for wanting to do the Ph.D. Ideally a reason that would require you to go beyond an MS.

• The ability to see a project through to completion and to understand that process

Ph.D’s are all about independent research. For many students, they have been told what to do for the entire time they’ve been in school. Working independently can be really scary, but these growing pains have to be overcome. A publication is evidence that you can do independent research, but it’s not everything. Even if a project is only part way finished, you can demonstrate maturity and depth if you can talk about all the motivations and reasons behind the choices you make for it.

Case in point: I got through 3 projects while I was in my MS. None of them were published at the time I started applying to Ph.D programs, but I did everything I could to prove that I could be independent and hard working. I guess one school took the bait.

Hope this helps someone. See you all in the next one.

Christian

---

A Snapshot of Very Normal

😵‍💫 What am I working on right now?

• Editing the second part to the nonparametrics explainer

🧐 What am I enjoying right now?

• Book — I’m currently listening to is Uncommon Sense Teaching: Practical Insights in Brain Science to Help Students Learn by Barbara Oakley and Beth Rogowsky. I feel like I’m reaching my limits as a teacher, so I’m trying to learn more about how to better teach others.

• Thing — I got a camera! You’ll be seeing my ugly mug a bit more, not just on YouTube but in other pieces of content I have in the works.

📺 What are my recent videos?

• How we use randomness to improve our lives: a video on how random treatment assignment let us make the leap from correlation to causation.

📦 Other stuff

• I wrote guided solutions to problems to Andrew Gelman’s Bayesian Data Analysis. It’s for advanced self-learners teaching themselves Bayesian statistics

Heads up! Some of the links here are affiliate links, so I may get a small amount of money if you buy something from them. I only link stuff I actually use.