Why I'm Spending the Next 5-7 Years in Grad School

Written on April 24, 2017

It’s hard to justify spending 5-7 years in grad school instead of getting a normal job. and I am often asked why I made this choice. A friend thought I was crazy to not call Google back for the second interview. I hope this article will help others going through the same choice, and to help my friends understand why I’m so excited! :) Also, I look forward to reading this article in five or so years, to see how well my expectations held up.

I Want to Be an Expert in an Unusual Area

At the most basic level, a PhD is a question., the point of a PhD is to think of a question, read a lot to see if anyone else has already answered it, but if not, discover the answer yourself. The main reason I made this choice was that no where else will pay me a full time salary to study almost whatever I want. No job that I know of will give an amateur this much freedom. As I work to become a professional researcher, I will also benefit from the advice of professors and more senior grad students, experts in both the content areas and the research methods I want to master.

In my case, I want to understand why computer programming is so hard, and what kind of lived experience makes this process harder or easier. I am interested in how people think: how they learn, how they collaborate, and how they solve computer programming problems. I want to apply the things I learn to help improve the quality and equality of Computer Science Education. This best fits into the computer science fields of Software Engineering, Human Computer Interaction.

Would I learn these things working at Google? Maybe, but unless I get really lucky, I could only do so indirectly. Most fresh college grads work at Google as Software Engineers, and write the code that makes Google work. Doing this, I would undoubtedly develop some intuition into how people think while programming, but I wouldn’t have the opportunity to study this rigorously.

I Want to Develop Specific Skills

I have learned many life skills in undergrad, but not enough. Here are things I want to learn in grad school:

  • Self directed research: I want to learn how to work on long term, largely self directed projects while staying motivated and on task. I want to learn how to form questions well, and how to design methodologies to answer them.
  • Effective oral communication: I want to be able to give talks that make people go “wow”. Grad school will give me plenty of opportunities to give talks on things I care about, so that I can develop this skill.
  • Writing down highly technical concepts concisely: courtesy of the Honors College, I’m pretty happy with my ability to write well. But because of my liberal arts background, I’m far more comfortable writing about Shakespeare than Software Engineering.
  • Collaborative problem solving: I want to get better at solving hard problems in groups, under time pressure.
  • CS skills: I still get to take classes! I don’t know how Machine Learning works, and I’ve forgotten most of the Algorithms I used to know. In grad school, I’ll learn and relearn each, respectively.

I Want to Live a Certain Lifestyle

I’ve heard people tell me “As long as you keep your advisor happy, you can pretty much do things the way you want.” This means I can work a ton some weeks, and then take a vacation. I can get in early and leave early to hit the gym, or I can sleep in and work late. I grew up traveling, and I want this to continue: when your research paper gets accepted, your university will usually fly you out to present your research wherever in the world the conference is being held, with a few days at the end to explore the city.

Grad school isn’t all roses: you work a ton. Before committing, I informally polled 6 PhD students. One worked 45 hrs a week, four worked about 60 hours a week, and one said they worked 70-80 hours a week. However, they worked this much because they wanted to: they cared about their topics, they wanted to finish their PhD early, or they wanted to take long vacations.

phd comics you can work whenever you want Yay!? Taken from PhDComcis.com

Would Google let me live like this? Not really, you generally have less control over your working hours, and often you work a ton of hours but not really by your own choosing. And you don’t get to travel as much. However, Google would pay me far more money. CMU will pay me about $30k/yr and pay my tuition, Google’s Software Engineer salaries start at >$90k/yr. Even when you adjust for the large cost of living difference between Pittsburgh and San Francisco, this is still stacks more green. But, I still feel lucky to be paid anything to study what I want: many people have to pay for their grad school.

Bonus: Why I chose CMU

CMU is #1 ranked, but that’s not a good enough reason to choose a grad school. Here are some good reasons: who you’ll be working with (most important), the city where you’ll live, and the kind of opportunities that living in that city will provide. CMU has a huge faculty of computer science. The number of CS professors at CMU is bigger than the total number of professors in any area at other schools. This means there will almost certainly be someone who knows the answer to any CS question I could come up with, or at least one who could point me in the right direction.

More specifically, CMU has the Institute for Software Research, with many faculty working in the general research area I’m interested in, and certainly proficient in the methods which will help me learn what I want to learn. There are also a few professors who do research in the exact area I’m interested in, with whom I would love to work!

I had a great time in Pittsburgh last summer, and I know I will be happy there. Additionally, largely because of CMU’s presence, there are many big name tech companies in Pittsburgh, and its common for PhD students to intern there over summers and possibly even work there after graduation. So in my eyes, even if research turns out not to be my cup of coffee, I’ll have enough opportunities to get a “real job” after I graduate.

Note for grad students and professors: I’m sure I’ve been naïve and overly optimistic in at least some places above. If you want, I’d love it if you leave a comment or email me to help burst this bubble sooner rather than later!