Part II: Thriving in a graduate program in statistics

(Note: This is the second in a series about graduate life in statistics. For links to all articles in the series, click here).

Here are the best pieces of advice that I can give someone currently involved in a biostatistics or statistics graduate program.

1- Know your interests, and exploit them

Did you sit through a martingale theory lecture and talk excitedly to the professor afterwards? Or, instead, did you start to think you that you would have been better off as an actuary?

These type of gut feelings are useful when it comes to one of the most stressful periods of a graduate student’s career – picking a research topic. I’ve often been asked how I picked my dissertation adviser, but I’m not sure its a great idea to just pick an adviser before you’ve picked a topic. My best advice is to (i) find the type of methods papers in statistics journals that you actually enjoy reading, (ii) find a faculty member capable of leading a thesis or dissertation in this subject, and (iii) hope that it will be a good fit. Thankfully, this meant that I could avoid martingale theory, and alternatively, I found a niche in the field of causal inference.

Of course, there are several great ways of finding a research topic and a faculty adviser – there’s even a song about it – but at the end of the day, nothing is more important than you actually enjoying the reading, writing, and advancing of your area of expertise.  This is obvious, but if you enjoy your work, then it won’t feel like work.

And if you are worried about which faculty member you are going to work with?  Just go to their CV’s and invest a few moments in their research. If you find that one of their prior manuscripts makes you think critically about the topic, its likely that this faculty member still has relevant ideas that could lead into a nice collaboration. In general, a faculty member is going to have a difficult time saying no to a student who is excited about his or her research area and has done the requisite background reading.

Choosing a dissertation topic has led many a student astray. If you pick a topic that you love, as opposed to one that just your adviser loves, that won’t be you.

2- Always say yes

There are so many opportunities in graduate programs and their universities for students to meet faculty, collaborate with their peers, and maximize their time spent. Mostly, it just comes down to saying yes.

These types of collaborations could come with several different types of people, including PhD students in other fields, masters students (both in statistics and elsewhere), or faculty looking for some level of assistance. Projects can also range in both topics and depth, from sample size calculations, to editing a peer’s thesis, to attending and taking part in a talk, or to revising a manuscript for statistical clarity. In addition, most institutions have certificate programs, software courses, or teaching opportunities outside the department that are worth looking into. Statisticians can dig their hands into almost any researchers play area – so be ready to get dirty!

There are two benefits to putting yourself out there. First, there’s a reasonable chance that a few of these opportunities turn into interesting and provocative collaborations. Some projects, as small as they might seem at the beginning, might turn into manuscripts that are submitted for publication, and on a few of these, you might play enough of a role to be named a co-authors. These consulting opportunities are great ways to lengthen a CV.

Second, even if there is no formal manuscript submission, the opportunity to network could help down the line. When it comes time for the job application process, both formal and informal recommendations could play an important role. Further, in most instances, you will learn some fun research that other folks are working on, while hopefully having a chance to impart a bit of statistical wisdom. It never gets old explaining that there’s no important difference between a pvalue of 0.049 and 0.051.

Finally, the best, underrated aspect of saying ‘yes’ is that many collaboration opportunities often beget more collaboration opportunities down the road.

3- Stay active

Here are a few ways that statistics and biostatistics students can stay active in areas beyond the focus spots of their research.

Conferences: Find conferences, both near and far, and make every effort to get to two or three per year. While the travel can be expensive, registration will always be cheaper and reasonably priced for graduate students. Further, many schools, programs, or advisers can find a bit of money somewhere for graduate student travel.

Conferences are a great opportunity to meet both people in your field, fellow students going through the same grind as you are, and to find topics in statistics outside your research area. JSM, ENAR, and JMM are three of the main conferences to be familiar with, although specific areas of research will also have their own events.

If you can afford to attend, attend.

Social media: While I admit that this recommendation is a bit of selection bias – it is not totally fair for a student that loves social media to advocate for social media – I feel strongly that students interested in statistics should be avid readers and users of social media.

For starters, data journalism has become a really, really, neat field to be involved in, as companies like ESPN (via FiveThirtyEight), Deadspin, Facebook, and the Washington Post have all backed analytics-based research in the past few years. Moreover, even PornHub (don’t worry, the link is safe) has done some analytics research! I, um, found out about this site through a friend. Such research can be both entertaining and educational, and most people in statistics, myself included, could take a few cues from the data visualization aspect tools that these websites employ.

In addition, there are several other outlets to keep up with. Folks on twitter will live-tweet academic conferences, so even if you can’t attend your favorite meeting, you can still keep up with the action. Columbia’s Andrew Gelman posts nearly every day on his blog, which is a must read for anyone involved in statistics. Johns Hopkins, Minnesota, and Harvard School of Public Health are among the many programs with active twitter accounts that link to student activity and research development. And if you have something to say? Blogging is free, and within a few days, your opinion, graph, or research can be accessed instantly by anyone around the world. One of my favorite things about writing a blog, as opposed to an academic paper, is that the response is instantaneous, unfiltered, and not restricted to only two reviewers!

Greg and I, as an example of a social media success story, became friends through twitter and through attendance at the New England Statistics Symposium (NESS). This eventually led to a Kaggle victory, a journal submission, and at least one really awkward conversation:

“Hi, I’m Mike.”

“Hi, I’m Greg’s wife. How do you guys know each other?”

“Oh…we met online.”

Speaking of Greg, in the next part of this series, we will explore “What I wish I had known when I started a graduate program in statistics


1 Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s