Skip to content

Latest commit

 

History

History
55 lines (39 loc) · 3.25 KB

why-phd.md

File metadata and controls

55 lines (39 loc) · 3.25 KB

The What and Why of a Ph.D.

Advising and working with Ph.D. students is the best part of my job. In fact, it is probably the main reason I chose to be a professor. Ph.D. students are my most central and essential collaborators. The two main goals of my work with Ph.D. students are (1) to produce great research and advance the world's knowledge, and (2) to advise and train the student to become independent, knowledgeable, critical, and capable researchers.

The main requirement for your Ph.D. is to be passionate about scientific inquiry, pursuing research ideas and questions, and being willing to develop the methodological depth and expertise to address these ideas in the most robust-yet-attainable manner.

This is also your Ph.D., so:

  • You should have significant freedom and agency.
  • You should feel ownership over your research.
  • You should be excited about the research you do.

By contrast, I don't believe the Ph.D. is:

  • A "long class"
  • A program where I hand you a list of instructions for you to carry out
  • A way for me to extract labor to grow my CV
  • A way for you to boost your CV to get a job of (for any X, e.g., "professor")

Toward these ends, I will aim to give you as much independence as I think you can handle (and maybe a little more), as early as you can and want to have it. I try to assess this case-by-case; if you feel like I misjudged it any time (e.g., too much or too little), please let me know. The general path is increasing independence with increasing time.

How long is a Ph.D.?

I believe that your Ph.D. is over when you:

  1. Have satisfied the minimum requirements of the program
  2. Have made significant academic contributions
  3. Can get the job you want.

While 1. is clear, 2. and 3. need more explaining.

  • For 2.: What counts as significant academic contributions? Usually, one determines this via your publication and conference paper history. One paper with only minor community impact, published in a tier-two journal, does not meet the mark of significant academic contributions, to put a lower bound on things. At the same time, a single, very high-quality, impactful paper can be sufficient for a Ph.D. There are historic cases (often in math), where a single paper is copy-pasted into the Ph.D. thesis, and completing that paper constitutes everything one needs for a Ph.D. in the way of academic contributions. These are not the norm and often happen more by accident than anything else, so you cannot count on this happening. In doing research, I describe ways to balance major and minor expected contributions.

  • For 3.: What job do you want? I cannot answer this for you, but graduating with a Ph.D. in Computer Science, Engineering, or Physics from a top university opens many doors. The usual "triangle" argument goes as

    • Industry: Most pay, least independence
    • National laboratories: Medium pay, medium independence
    • Academia: Least pay, most independence. On a personal note, I heavily value the latter, and its often easier to hop from academia to industry than vice-versa. So, here I am.

The Ph.D. takes very roughly 4-7 years, with students headed to the academic job market lean on the longer end, say 5-6. However, your work ethic, project(s), and more can shorten or lengthen this.