The goal of the one branch ahead series is to demystify that leap we all must take from grad school to the great unknown that lies beyond. While the careers profiled here may inspire goals and illuminate future possibilities, there remains a gap between here (grad student) and there (successful professional) and with it, a question: How do you make the transition from grad school to a nascent career?
This installment of the series features answers graciously provided over email by a friend who recently finished his PhD in physics/computational neuroscience. Read on to learn how he became a statistical epidemiologist, why he turned down post-doc in Paris (!), what he wished he had learned during grad school, and where his work has an impact now. Be sure to read the last question, where he gives advice for current students that is both widely applicable to grad students and particularly useful for those with technical/statistical aspirations.
What did you study in your PhD program?
My PhD is in Physics, but my thesis area was computational neuroscience: math and biophysics of single neuron computation. Lots of time with pencil, paper, and computer.
When you first started the program, did you know what you wanted to do after grad school?
I really never thought that far ahead back then. In undergrad, the major I ended up sticking with was Physics Education – I trained to be a high school physics teacher. I taught for a year after graduating, but realized quickly that I wasn’t done with my formal education. I went to grad school to learn more, without really thinking about career implications beyond that a PhD in Physics wouldn’t hurt.
How did your career plans or goals change by the time you were finishing your PhD?
For most of grad school, I figured I was on the liberal arts college professor track. Teach first, research second. This plan involved surviving the very strong urge to drop out with my masters during the fourth year because I wasn’t happy as a researcher (too much time working alone).
I stayed in, thankfully, by the following logic:”It’s July. By the time I wrap up, it’ll be September and I’ll have missed the hiring period for high school teaching. So I can’t get work until January, but that’s ski season and I won’t want to leave. So I should stay until May. I’ll have been here almost 5 years then. I should really get my PhD after being here that long.” When I was getting ready to get ready to graduate, it became clear that it’s hard to get teaching professor positions without a successful research post-doc, and that most of the colleges I would’ve liked as a professor aren’t in places I’d want to live.
The plan changed to figuring out how to facilitate an interesting life while doing something tolerable for work. I was able to set up a post-doc in Paris, but while writing fellowship applications, I had to accept that I had absolutely no room in my soul for more of what I’d been working on and that not even Paris was worth staying in my field.
So I walked away from the post-doc, but stayed on the new plan.
What do you do now?
I’m an epidemiologist. I do math modeling and data analysis in support of the Global Polio Eradication Initiative. My title is Research Scientist and I’m employed in a not-for-profit division of a corporate research lab belonging to an infamously for-profit company. My group works on polio, malaria, and HIV, and TB in relation to HIV. Most of the research I do could be called genetics and virology, which is funny because I hadn’t studied anything remotely related since 9th grade.
What’s a typical day for you like now?
Most of my day is similar to how it would be in a large academic group. My team has a dozen researchers, and we all spend a lot of time in the weeds with data, coding, debugging, building and checking models… There are science and organizational meetings once a week, and I usually get to research full-time Wednesday through Friday. I also work from home almost once a week, and that adds a lot to my quality of life.
About five percent of my time is spent maintaining external collaborations. We work with the WHO, CDC, Gates Foundation, and county partners in Africa and Asia. Essentially all of our research is motivated by the needs of our partners and are not determined by the taste and interests of the PI. We choose how to interpret those needs, but we have to be agile. It’s not uncommon for scientifically interesting projects to get put on hold or abandoned because something more relevant to global health policy has to get done.
The policy role means we spend a lot of time thinking carefully about language and presentation. My third most important output is a scientific paper. The first communication priority is to package quantitative science for less quantitative, and generally more powerful audiences. The second is to help transfer analyses that we develop to our partners. The ideal project for us is to develop something useful that is simple enough for our partners on the ground to do themselves.
If you’re wondering why I am happy now as a researcher while I wasn’t in grad school, it’s for two reasons. This group is much, much more collaborative while keeping project ownership well-defined. I love this.
The big picture is if I do my job well, through the scientific process and elaborate bureaucracy, there will be fewer paralyzed children.
Were you involved in any organizations or activities while in grad school that helped lead you to where you are now?
The communications skills from all the teaching experience before and during grad school are definitely relevant. Maybe less obvious was that I often went to seminars outside my department. The connection that got my resume in the right place came from a professor in Applied Math who had a good opinion of me even though we never really worked together. The more people you are acquainted with, the more likely one will know of something other than academic jobs.
Of the skills you developed as a grad student, which are most useful to you in your current position? Which are you happy to leave behind?
Most useful: “No, I don’t know anything about that, but I bet we could figure it out.” And: “I don’t think that number means what you think it means.”
Broadly interpreted, I don’t think I’ve left any skills behind.
Are there any skills/tools/habits that you need in your field but didn’t develop during your grad school training?
I would be more productive now for the same amount of effort if I was a better programmer. In grad school, I did everything in Matlab, and because I worked alone most of the time, I did everything sloppily. I’ve gotten a lot better since, but I’m still mostly locked into Matlab. Every week, there is something I do that would be faster or easier in C, R, or Python if only I didn’t have to pay the start-up cost.
What’s the best thing about your job?
I can sincerely say that everyone I work with closely is really, really smart, and they are more cooperative than competitive. Most of my colleagues are past the post-doc phases of their careers and could be tenure track faculty if they wanted that life. I’m one of the least experienced people in the group.
Where do you see yourself going from here?
I hope to have this job for at least the next 4 years so I can finish paying off my loans from undergrad. It’s relevant that I now expect this to take 4 years and not 15. I could imagine happily following this career path for much longer, but given my wandering interests and the inherent instability of research support in the private sector, it’s important that I keep my eyes open.
Any words of wisdom for current grad students who might be interested in making a transition like yours?
Do what you can to become broadly knowledgeable about your science (and not just your field). The more things you know at an end-of-first-year graduate student level, the more doors will be open to you. Because I was an unfocused researcher in grad school and am a generalist in my soul, I probably spent 20% of my working time in school learning stuff that wasn’t obviously related to my research area. That percent may have been unreasonable, but my breadth (in my case, statistical inference, stochastic modeling, and a solid grounding in the math behind most of physics) is directly responsible for me surviving my interviews and succeeding now.
Getting a post-doc required talking over coffee a few times for a few hours, giving a well-received talk, and having an advisor who could speak for you. Getting a technical position in a company that arose from software culture required surviving 10 hours of oral examination over three interviews, often on subjects that, when defined narrowly, I never thought about before in my life.
Even the first interview, nominally an informational, lasted almost 90 minutes and involved a question about modeling doctors as disease vectors in a hospital. Even at the most cognizant companies, the interview process is biased in favor of talky alpha-types, if only because they give out information about themselves faster. So, to be ready for technical interviews, try to become comfortable thinking aloud even if you have no idea if you’re on the right track. When given a problem to work through, ask lots of questions about it. Talk through why you’re making the assumptions you’re making. Talk through what you might be screwing up. Talk through what you’d need to know but don’t. Pay attention to their body language if they aren’t giving a lot of feedback. And, if you need a minute to think quietly, say you need a minute and take it.
Learn to program decently if you think you’ll want a technical position. Even if everyone you know uses Excel and you think I’m crazy. I’m not saying you need to be a software developer, but basic skill and familiarity go a long way. And, if you end up in a job where data are looked at but programming isn’t common, it’ll be as if you are a wizard among mortals.
Be competent for your needs in one programming environment. Almost anything you can do in Excel or SPSS, you can do faster, more flexibly, and more reliably in R, Matlab, or Python, and there’s a million things you could do that you can’t without programming. You can even write Excel (or whatever) output automatically for your luddite boss after you’re done. Be able to do in another language the easy things you do in your good language. You’ll be more conversant in interviews, and you’ll be much more likely to know when a new problem will be easier to solve with your non-preferred tools.