The Ph.D. Grind - Answers to Reader Questions
September 2012 (postdoc)
Question: You mentioned that conferences are more prestigious than journals in computer science. However, influential people in my university value journals more. I'm confused; who's correct? Also, how reputable are certain international journals?
The short answer is that prestige is relative: If you're in a university that values journals, then you need to publish in journals to advance your career. But in top-tier U.S. academia (circa 2012), computer science conferences are definitely more prestigious than journals. Read the following description of CS conferences versus journals from Professor Michael Ernst. Also, there are plenty of terrible journals (and conferences too) around the world that accept just about any submitted paper; see the SCIgen stunt for more details.
Question: If my goal is to enter into industry (e.g., as a scientist or engineer) rather than becoming a professor, then should I still pursue a Ph.D.? If so, how should I go about applying?
It depends. Some employers view earning a Ph.D. as evidence of intellectual maturity, personal responsibility, and ability to follow through on difficult projects. Thus, they might be willing to hire you into more interesting or challenging roles, even if they don't care about your specific Ph.D. dissertation topic. On the other hand, other employers might feel that someone fresh out of college is a better fit for the jobs that you want.
However, if you're not from the U.S. and want to work in a U.S. science/technology company, then getting a Ph.D. from a U.S. university will likely be beneficial. You stand a better chance of getting hired by a U.S. company with a Ph.D. from a U.S. university than with an equivalent (or lower) degree from a foreign university; in general, U.S. companies "trust" U.S. university degrees more than foreign degrees.
For application tips, read my advice for Ph.D. program applications, and pay close attention to the final section, "Do you have any specific tips for non-American students applying to American universities?" In short, apply to a broad range of schools to maximize your chances of acceptance; don't just limit yourself to the top tier.
Question: I'm very interested in doing research in field X. Can you give me advice for how to get started? For instance, how should I find out which papers to read or how much to read?
First, ask yourself why you want to do research in field X. Is it just for your own personal intellectual curiosity? Or is it to launch a paid career in research? If your goal is to start a research career in X, the only way to get started is to find a professor or researcher in your local university who is interested in field X and to start working with him/her on it. For those without the privilege of such access, I'm afraid that it will be very hard for you to enter the research community. Sure, you can do research for its own sake or for personal fulfillment, but without institutional support and mentorship, it's nearly impossible to start a research career. If you're one of the rare people who can do it, then you don't need to be reading this article! Also, see my reflections on Ph.D. student life for more details.
Question: How can something still count as your own research if your advisor helped to think of the idea and lays out the direction of the project?
It's common for undergraduates, masters students, and even some Ph.D. students to be handed an already-formed research idea by their advisor. The student's contribution is figuring out and implementing the details to get these initial research ideas into a working and publishable state, which is often non-trivial.
Question: How can I implement my own creative research ideas rather than just working on projects that my advisor hands to me?
I wrote an article about this topic: Lead From Below
Question: I was impressed by how you were able to recall the details of your prior research projects when writing The Ph.D. Grind. Do you have any suggestions for writing up and organizing ideas, especially half-baked, vague ones that are common in research?
Here is the super-simple notetaking system that has served me well throughout grad school (and beyond). I keep a single text file called my "work log" and append new entries to it every day as I'm working. At the beginning of each day, I simply write the current date as the heading and then jot down whatever notes I feel like jotting down. Each day's work log entry ranges from a few bullet points to a few paragraphs, maybe with copied-and-pasted snippets from email exchanges or code samples. I keep everything very informal and don't bother organizing at all. Then every few months or so, I go back and re-read my old log entries; inevitably old ideas mix with new ones to spark some novel inspiration. In addition, I keep a separate to-do list of work-related tasks for each project. Read "The Spark File" by well-known technology writer Steven B. Johnson for a more eloquent description of this simple technique.
Question: How did you recall your emotions in so much detail when writing The Ph.D. Grind?
I pored through tons of old emails that I exchanged with my advisor, colleagues, and friends throughout my Ph.D. years; those emails were very effective memory triggers. In addition, I also read through my old research notes (see previous question), paper submission drafts, and feedback from paper reviewers to further jog my memories.
Question: Which classes did you take throughout your Ph.D.? Which classes should I try to take?
The short answer is that I took as few classes as possible (3 or 4, I don't even remember). One reason why I never mention classes throughout The Ph.D. Grind is that the primary purpose of a Ph.D. is not to take a bunch of classes; it's to produce research output so that you can eventually graduate. In my limited experience, there is an inverse correlation between emphasis on classes and research productivity; that is, students who are focused on (and love) classes are usually the ones who have trouble graduating on time. So my advice for taking classes is to only take whatever you need to make yourself more effective at producing research in your chosen sub-field. I know this is an extreme stance and have already heard the counterarguments, but I strongly stand by it.
Question: How did you try to remain productive day-to-day throughout grad school when there was so much uncertainty in the work?
I've written up my thoughts in an article entitled How To Be Effective.
Question: You did three summer internships throughout your Ph.D. That seems like a lot. Do you recommend Ph.D. students to do internships?
Absolutely. I recommend all Ph.D. students to spend at least one summer, preferably two, away from school working at an internship. (Three is a bit extreme, though; my third internship was unplanned.) In general, professors don't like their students leaving for the summer, since they want as many students in lab "producing" for them ("producing" is the exact word used by one professor I know). But I strongly insist that you insist on spending at least one or two summers away from campus. At worse, you delay graduation by six months, but the real-world industry experience and connections you gain more than make up for that minor delay. When you graduate, you won't be guaranteed a decent job in academia, so you will need to look for industry jobs; having internship experience gives you an advantage over fellow Ph.D. grads who spent all of their summers producing for their advisors.
Question: Can I translate The Ph.D. Grind into another language?
Since I cannot fluently read any other language besides English, I cannot assess the quality of translations. However, if you are deeply passionate about translating, please first send me an email asking for permission. In general, I might allow translations as long as three conditions are met:
Last modified: 2013-03-01