What You Can Expect

Everyone can expect the lab to be a positive, inclusive, and collaborative environment. That said, we each have the responsibilitiy of committing ourselves to creating, and maintaining, just such an environment. The lab is the people in it.

What is grad school?

I sometimes repeat myself; I sometimes repeat myelf (sorry). One of the mantras I repeat over and over to students who want to get involved in research is that grad school is completely different than undergrad. How is it different? The following excerpts from a Twitter thread started by Dr. Chanda Prescod-Weinstein address exactly this question. I find all of the comments below not only to be spot on, but also very practical in terms of providing actionable advice. Thanks to Paul Gribble of Western University for posting this in his lab manual.

The hardest phase transition for PhD students, I think, is that you are supposed to be working toward competent intellectual independence. That means that your advisor doesn’t solve the problem: you do. You write your grant and fellowship proposals.

And in the case of the actual science: it’s not a problem set. There are often not step by step instructions for how to solve the problem. You are finding the instructions and implementing them and sometimes your first through fourth tries fail and it’s no one’s fault.

If working at the boundaries of what we know — being confused often and working hard to reduce the confusion — doesn’t sound like fun work, research may not be a fit for you, no matter how well you are being compensated.

Of course, as in all professions, sometimes your boss [advisor] is an asshole. But sometimes it’s just hard and not because you’re being abused.

Research is mostly the part where you are doing the work, not mostly the part where you are presenting known results. If you mostly enjoy presenting neatly packaged results, that’s a different kind of problem solving: scicomm is great and challenging and different from research.

Research requires a kind of confidence in your own logic, that you can reason your way to a solution because the solutions are a matter of reasoning. So it requires a flexible intelligence mindset: you can learn the techniques you need to know.

But no one is going to do the learning or hard work for you. Nor should they. If your advisor does too much for you, you never become competent to do things on your own. It is a difficult balancing act that is hard to get right.

One of the most important roles of faculty advisors: helping you pick the right problems; topics that are unsolved, do-able in ~3 years, will be of interest to the community and within the student’s skill set.

What is graduate research about?

Along the same lines, I found some great advice about graduate research on Sam Gershman’s lab website. I’ve copied some of my favorite points from his slides below. They all resonate with me, and I definitely recommend everyone go through the entire pdf (found here) themselves. The added emphasis to some sentences is mine.

Many students struggle to find an appropriate research project when they start grad school. I think this stems from the fact that science is taught as a body of knowledge rather than as a body of epistemic activities. So undergrads leave college “interested” in certain topics, but with little idea about what one does with those interests. Science is a puzzle-solving activity.

In an influential paper, Platt (1964) asked why some fields have made faster progress than others. His answer was that the successful fields made use of strong inference: the design of experiments to discriminate alternative hypotheses. This seems intuitive and obvious, but a lot of science is done without theories that make clear and distinctive predictions. Many scientists seem to have the attitude of “let’s measure stuff, and then from these measurements we can construct a theory.” In my view this random walk in the space of experiments is hopeless.

I encourage my students to work on two timescales at the same time: at short timescales (months), work on small puzzles, while making progress on a large puzzle at longer timescales (years). The psychological impact of having produced intellectual output, no matter how small, is not to be underestimated. Multiple timescales also help alleviate boredom and frustration.

In other words, grad school and research are largely about finding your way when it comes to solving interesting problems. You build your own curriculum to help you learn how to identify and solve puzzles, both big and small. Solving theory-driven, thematically-linked puzzles that align with the lab’s goals is how you build your thesis/dissertation.

Reading papers

When I was a postdoc, a labmate once shared this gem with me: “A month in the laboratory can often save an hour in the library.” Hilarious, yet true. We can all thank American chemist Frank Westheimer for this one.

The point is, you must read papers—lots of them! This is how you familiarize yourself with the relevant scientific literature. It’s also how you develop questions, important ones. And it’s impossible to have a deep understanding of your research area without extensive reading. Reading is also a great way to learn how to write. I don’t know of a single good writer, scientist or not, who is not also a voracious reader. The Onboarding section has some suggestions on where to begin. I can also help guide you.

Start your reading journey by choosing a reference manager like Zotero, Paperpile, or JabRef. I used to use Zotero and found it very intuitive but have recently switched to JabRef, only because I like its easy integration with \(\LaTeX\). Choose whichever one works well for you.

N.B.: No matter how important reading is (and it obviously is), don’t lose sight of the fact that “Science is a puzzle solving activity” (see above). You don’t have to read X number of papers before you get to do science, and reading papers and doing science don’t have to be seen as separate, either. Yes, you must read papers to have a good understanding of the field, and before you embark on a big project, you should know what’s been done in that area already. However, you can (and should) turn your reading into science (i.e., puzzle solving): Read, ask questions, and then try to answer your questions (usually through some combination of sketching on the whiteboard, math, simulation, and/or an experiment), and repeat!

Making progress

Your progress, at least when it comes to external markers, will be highly non-linear. Some days, or weeks, you’ll be making great “progress”, whether that means a paper or grant gets accepted, you pass your quals, or an experiment produces the predicted results. Other days/weeks/months, these things won’t happen. Much of this is out of our control. What is under our control, however, is the amount of focus and effort we put into our work. If you dedicate a good deal of both, you are virtually guaranteed to make significant progress towards being a better scientist. And since that is what you care about, your work will by definition be meaningful.

That said, come speak with me if you are unsure of what you are doing, or what you think you should be doing, on a day-to-day basis. We all go through periods of doubt and uncertainty. I am here to provide guidance and help you through those times. Your lab mates are also a source of support.

Concretely, a good strategy is to try to carry on with your reading while actively seeking out answers to questions. That is, make sure your didactic work goes hand-in-hand with your active learning and exploration. It’s never too early to work with a model, design an experiment, re-analyze some data, or attempt to replicate a previous result! Throughout your time here, put yourself in the mode of doing science.