Philip Guo (Phil Guo, Philip J. Guo, Philip Jia Guo, pgbovine)

My thoughts on "The N=1 guide to grad school" by Adam Marcus

My friend Adam Marcus wrote a wonderful retrospective on his years as a Ph.D. student in the MIT Computer Science and AI Lab, called The N=1 guide to grad school (and hopefully, knowledge work).

It's really good. I nodded my head so many times when reading it that my neck now hurts.

Here's what went through my mind as I was reading:

As Adam mentions up-front, everybody's Ph.D. experience differs based on factors such as field of study, school/advisor reputation, and—the F-word—funding. To combat this pervasive “N=1 problem,” I wish that every graduating Ph.D. student would write a similar retrospective immediately after completing their dissertation. In a similar vein, a few years ago, I pitched the idea of conducting “exit interviews” with each graduating student in my department and then compiling the responses in a guidebook for incoming students. My friends feared that such a guide would decrease enrollments, so I never pushed on the idea.

Adam's advice to be positive is awesome! (Yes, I was being positive.) I'm a naturally positive person, so for me, the most awkward encounters in grad school were meetings where expressing negativity was the way to show intellectual strength. For instance, during my first few years of grad school, I felt uncomfortable speaking up during group meetings and paper reading discussions ... not because I felt incompetent, but because I felt like I didn't have the proper negative/scathing/caustic remark that would make me sound as smart as the senior students and professors. Of course, these fears seem silly in hindsight. I think that the most senior person in the room (whether a grad student, postdoc, or professor) should lead by example, demonstrating that it's possible to be critical and intellectually rigorous while remaining positive.

“Your advisors are smart collaborators with lots of experience, but they aren't your boss, and have no training in management.” What Adam mentions in the last part of this sentence is likely a large contributor to grad student woes. Your advisor was hired based on their accomplishments and potential as a producer of research (papers and funded grants), not on their accomplishments and potential as a manager of people. In the ideal case, being a good manager leads to good research output. But I've also witnessed professors churning through student labor to get the results needed for tenure, promotion, fame, and continued funding, leaving behind a mess of burned-out and resentful ex-students. I wonder how advisor-student dynamics might change if a non-trivial part of a professor's tenure or promotion case involves confidential interviews with all of their current and former students?

Another thought I had on Adam's You're in charge section is that by fully taking charge of your own Ph.D. journey (within the constraints of your funding stipulations), you make it more likely that professors will want to help you. Paradoxically, the people who get the most effective help are those who demonstrate that they can best help themselves. As I wrote in The Ph.D. Grind: Lead From Below:

... imagine a professor with two Ph.D. students: Alice and Bob. Alice comes into each weekly meeting with an optimistic yet realistic tone, summarizing what she tried in the past week, honestly describing what worked and didn't work, proposing a plan of action for the upcoming week, and then asking politely for feedback. In contrast, Bob comes into each meeting sounding cynical and frustrated, blaming last week's failures on colleagues, and whining about how he is being treated unfairly by everyone around him.

Which student is the professor looking forward to seeing each week? Who is the professor more willing to leave alone to continue their planned course of action? And who is the professor more willing to help?

Think hard about whether you're more like Alice or Bob. Does your boss look forward to or dread meeting with you every week? Do they smile or sigh when an email from you enters their inbox? When they talk about you to their colleagues, do they say positive or negative things? Are you a pleasure or a burden to manage?

Adam's section on Creating things meshes well with my work philosophy of “keep making stuff.” It's a must-read for anyone in a creative field. Ironically, even though people think of the scientific process as “creative,” Ph.D. students in pure-science fields might find it hard to follow some of this advice, since their day-to-day grinds involve analyzing existing phenomena rather than creating new artifacts.

I'll push back a bit on the concept of “crappy first drafts.” Although I totally agree that getting out a crappy first draft is better than producing nothing, there are a few dangers to be aware of:

  • You might be tempted to keep iterating to incrementally improve your crappy first draft rather than throwing it away entirely and starting from scratch.

  • Crappy first drafts might not instill confidence in others, so your colleagues are less likely to get excited about your ideas. First impressions and polish matter.

  • Crappy first drafts might “anchor” expectations and comments at a low level of discourse and not lead to optimal feedback.

To combat those dangers, the best way to make a crappy first draft is to make it so obviously “crappy” that people don't have unrealistic expectations. For instance, whiteboard doodles and hastily-scribbled bullet point notes beat a nicely typeset document for disseminating early-stage written ideas. And paper prototypes and Photoshop mock-ups beat a semi-working, buggy software application.

“It also helps to figure out what your research community thinks is cool.” Yes! As soon as a grad student discovers the intersection of what they feel is cool and what the research community thinks is cool, then they can start making progress toward a fulfilling Ph.D. project.

“Schedule two to three hours of low-interruption time every day you plan to work. Block off time on a calendar if you have to.” A hundred times yes! My productivity (and job satisfaction) directly correlates with whether I can get in three hours of focused grind time each morning before I start getting distracted by errands and obligations. Make a bit of forward progress every day. Don't break the chain.

The way I think about Adam's Today doesn't matter concept is that, over the course of an entire Ph.D., only the integral (total area under the curve) matters, not the height of each individual day's output.

I'm happy that Adam is also a fan of grad student side projects. See The Ph.D. Grind: Main Grinds and Side Grinds

I didn't follow nearly as much of Adam's advice in Putting yourself out there as I should have. Fortunately, by the end of grad school, I was respectable on the “selling my ideas” front. But in my early years, I felt very isolated and uncomfortable engaging with others about my research. Good advisors should foster an environment of inclusion and encourage their early-stage students to get over such fears as soon as possible.

Even though my advisor had plenty of funding, I felt guilty asking to attend conferences where I didn't have a published paper. I even felt guilty submitting workshop papers and using up funding to attend when those papers didn't “count” toward my dissertation. A part of me wishes that I had gotten more encouragement from professors in my department to attend conferences even when I wasn't a presenter; but in the end, I still blame myself for not taking enough initiative.

Hmmm ... “But unlike working at a company, where there are public relations and marketing folks whose job it is to make sure people know about you [...]” I would argue that at a company, PR and marketing folks make sure people know about the company's products, not about the individuals who toiled to make those products. Unless you're in a leadership or other public-facing role, it's hard to get publicly known for work you do at a company. One cool aspect of academia is that you're more easily able to get individual recognition for your work.

I would have benefited from working more with others throughout grad school. My Ph.D. was quite a solitary journey—I helped out on a few other people's projects but never maintained that sort of “pair research” flow that Adam speaks so fondly about. The closest I ever got was working with my officemate Peter on Klee-UC during my third year, but we were utterly demoralized after only two months. It was horrible. Maybe that trauma made me shy away from collaborative projects during my second half of grad school.

Adam also cautions against joining projects that you're not genuinely excited about: “But before committing yourself to a project, step away from someone else's excitement for it, and ask yourself whether the topic matters to you.” I'll push back a bit and assert that it's okay to join such projects, especially in your early years. Latching onto other people's excitement is a good way to learn more both about technical topics and also about the process of doing research. For instance, I never had a passion for string constraint solvers (gasp!), but I learned about that sub-field and, more importantly, about the mechanics of remote collaboration from working on the HAMPI project during my third year.

Finally, as Adam mentions in Take breaks, living with a significant other—especially one who is a “civilian” (i.e., not a Ph.D. student)—can help you stay sane throughout grad school. If you don't have a significant other, either find one (for the sake of your research!!!) or move in with civilian friends. Without outside perspective, it's too easy to get caught up in an echo chamber of mutually-reinforced grad student angst.

And now I'll take Adam's advice and just post this crappy first draft rather than spending too much time refining it :)

Created: 2013-05-08
Last modified: 2013-05-08
Related pages tagged as Ph.D. student life: