To PhD or not to PhD

September 5, 2006 2:12 PM

Should I go to graduate school to get my PhD in Computer Science?

Here is my situation.

I have a Bachelors degree in Molecular Biology from Pomona College.
I am on my way to finishing my Masters degree in Computer Science from Cal Poly Pomona. I have been working for the past 5 years as a software engineer at a small startup.

I'm debating whether I should go to graduate school for a PhD. I just turned 33 today and am having a midlife crisis with regards to my career. I feel that what I'm doing on a daily basis is not impactful enough. At least at my company, I always feel like I'm doing the bidding of the marketing and business people.

I'd like to hear about what graduate school in computer science is like. What time of routine would I have? What kind of stipend would I expect to have? What is a typical workload? What are my options after graduation? If I'm not really interested in teaching, would a PhD still be a good idea?

posted by zzztimbo to education (12 comments total)

I am not a PhD, nor a computer scientist. However, I believe that real-world real-job experience will get you farther than more schooling, unless you have a very specific goal you're shooting for.
posted by gnutron at 2:19 PM on September 5, 2006

It would seem that the idea here is to move outside the traditional career path.

At least at my company, I always feel like I'm doing the bidding of the marketing and business people.

That's business. period.
posted by bitdamaged at 3:20 PM on September 5, 2006

(CS PhD here.) Tell us what you want to do! What research interests do you have? What is your dream job?

Most people I know that have PhDs in CS or related have done:

a) professorship
b) company founder
c) research scientist at corporate lab
d) dev/cto position at startup or large company

Their stipends have been between $0 a month and $2500 a month. Depends on the school and grant and location. They have been terribly happy and terrible. Depends on the school and the students' decisions.

Please know, at least generally, what you want to do. Have a dream vision in mind, your ultimate dissertation title and your job title when you get out. The worst (stress and achievements-wise) grad students I've seen are the ones that applied and went out of some obligation or trajectory they thought they should be on. The happiest and best ones truly believed in their field and research. So from your field (I assume something to do with biology) find the top 5-10 schools and research programs, do your diligence on where you could live, who you like, etc. Make some introductions over email. Get some visits before you even apply. Try to talk to other grad students.

Workload, routine-- you'll work for about 4-5 years, on average. Take classes for the first two, do quals for the third year, propose right after that, then write+defend your 4th and 5th maybe. The classes will be challenging but not impossible. Some people make it a 9-5 thing, raise a family, go home at night and make dinner! Some work all night at the lab and eat food left over from meetings. You'll make a lot of friends. You'll be generally poor but not destitute, and after the 3rd year or so you'll stop caring about money so much. You'll travel a lot to conferences, deal with a lot of politics, get in a few fights with your advisor and admins, burn out once or twice. You'll become well known in a tiny subfield that 25 people care about. Maybe summers you can intern and make some money.

Re: "marketing and business people" -- that won't change. It's an attitude thing, not a credentials thing. You're going to have to become 50% business yourself to get over that. Someone always is paying for your salary, and unless you've invented an immediate profit-generating technology that has no capital expenses, it's not you. Getting signatures on a dissertation does not fix this problem.
posted by neustile at 4:14 PM on September 5, 2006 [1 favorite]

Yeah, your PHD won't change your standing w/ regards to Marketing people... if they push you around now, they'll push you around when you have your masters... They see a computer/biology geek, and they're obviously the ones who've been educated to know how things should be done.

However, if you start being assertive, make good decisions, and make sure others know that you made those decisions BEFORE the marketing people did, you'll be pushed around much less. You won't ever be able to push marketing/business people around, but if you are confident enough that you make good decisions, being assertive will help your reputation immensely.
posted by hatsix at 6:44 PM on September 5, 2006

Re: "marketing and business people" -- that won't change. It's an attitude thing, not a credentials thing. You're going to have to become 50% business yourself to get over that. Someone always is paying for your salary, and unless you've invented an immediate profit-generating technology that has no capital expenses, it's not you. Getting signatures on a dissertation does not fix this problem.

I second this. You always answer to someone in the end. Try reconsidering your job in light of this before considering a Ph.D. That way you won't leave work to escape something that you will find everywhere.
posted by halonine at 7:10 PM on September 5, 2006

I'm a Computational Linguistics/CS Ph D. Student.

What do you want to do with a CS Ph D? Do you want to do research? You will be equipped to work at research labs like Google, IBM as well as carry out your own research in an academic setting. A Ph D. is less about learning tons about one subject area and more about creating people who can investigate and solve problems independently.

Hours vary. It averages out to about 40 hours a week for me, but there are good weeks and weeks where something needs to be shipped. This probably resembles your current job. If you think you are going to have complete control over whatever you research you should probably consider that this will likely not be the case. Your funding will come from a project and if you are not insane your thesis will be a part of whatever brings in money. Want autonomy? Get your own fellowships.

Also consider that there is a good chance that you will be working on the same thing for years and years. I love my thesis topic and I am lucky to be building something that will help people, but sometimes it's like I've been eating my favorite meal for four years. Some days it's like "Hooray! Lobster again!" other days I feel like "Please let me eat something else."

Stipends in CS are generous, but can be tied to the cost of living in the area. Schools in Boston offer lots more money than ones in Pittsburgh. Either way, you will probably be given enough money to live comfortably. I managed to buy a house with my stipend.

Also, many CS schools will not make you take qualifying exams. This will save you some stress though you will have to take a minimum number of classes.
posted by Alison at 9:07 PM on September 5, 2006 [1 favorite]

Most of the advice here is pretty accurate, so I'll just try to supplement it.

I'm in grad school for CS. You will probably take classes for 1-2 years, which will be similar to your undergrad but possibly more interesting and less mundane. The second phase is doing your research and hopefully by then you have an advisor and topic, and your schedule is very flexible at this point. Some days you will read a lot, some days you will write a lot, and some days you'll just do laundry or something.

You may have up to 3 exams - prelims, quals, and defense. The defense isn't really an exam but an oral presentation in front of your committee.

My stipend is $1800/month this semester, which is plenty where I live. I spend $500/month on rent+utils+food and bank the rest. It may be higher if the cost of living where you are is higher, but don't expect it to be much more.
posted by tasty at 9:29 PM on September 5, 2006

"Stipends in CS are generous, but can be tied to the cost of living in the area. Schools in Boston offer lots more money than ones in Pittsburgh. Either way, you will probably be given enough money to live comfortably. I managed to buy a house with my stipend."

Maybe times have changed? Pittsburgh (CMU CSD) was offering a higher stipend last year than Boston (MIT CSAIL). Not much different, but still more.
posted by heresiarch at 6:53 AM on September 6, 2006

for the longest time i've been fascinated by biological influenced methods in computer science: swarm intelligence, ant colony optimization, emergent systems, neural networks, etc.

my ideal job would be as a CTO or a director of a research lab.

i love reading papers about topics that i'm interested in, but i often get intimidated/lost by the mathematical equations. i'm wondering if this would be a barrier or if it just requires discipline and training?

UC Irvine is very close to my house and would probably be my choice for schools. i have a lose connection to one of the researchers i may be interested in working with (he is on the advisory board for my company).

and the last question, am i too old to do this (33)? are there many disadvantages to pursuing a PhD at such an age?
posted by zzztimbo at 9:22 AM on September 6, 2006

"i often get intimidated/lost by the mathematical equations. i'm wondering if this would be a barrier or if it just requires discipline and training?"

I think it would just require discipline and training. My calculus prof stressed that the most important thing we should be doing in the course was developing our mathematical intuition, and by the end of the course I was a lot more comfortable with proofs and math in general. It sounds like you would benefit from something similar, so maybe a few theoretical (as opposed to the more applied sort you'd normally take for science) math courses would do the trick for you.
posted by benign at 9:44 AM on September 6, 2006 [1 favorite]

Don't sweat being 33. There are many people in my department in school after years working in the real world and a few are in their 40s. I think the jeans and t-shirts make everyone look younger so don't worry about being too old to fit in.
posted by Alison at 9:05 PM on September 6, 2006

On equations generally:

It seems to me if you're already getting a Master's in CS that you have the potential to understand equations. If you have too many emotions about them to spend time with them, that will be problematic. Granted, you may have a subtle learning or cognitive disability that will prevent you from being a whiz at them. If so, remember that genius is one percent inspiration and ninety-nine percent perspiration. You won't know either way unless you really start working with equations.

The topics you are interested in seem mathematics-heavy. You'll have to ask yourself if any possible struggle is worth it to you. Everyone's weak in something.

Ah, the Claremont Colleges! I almost went to one, once upon a time.
posted by halonine at 4:39 PM on September 7, 2006