Tuesday, November 10, 2009

Graduate School? How to Decide...

What do people think about students going to work for a year or two and then applying to graduate school? Or applying but then deferring to work for a year or two?

It's that time of year when seniors are thinking about graduate school. (I have multiple requests for NSF letters pending...) So, naturally, the other day I talked with a student who, essentially, had the question, "Should I go to graduate school?"

In this case, the question wasn't one of talent; the student would, I'm sure, do very well in graduate school. But he also has a job offer from a top company in computing where he could do interesting work and, I'm sure, also do very well.

In these tough situations, I try my best not to give direct advice, but instead try to get a student to talk about their own concerns and issues to help them realize which way they really want to go. While I feel positive about the outcomes from my having gone to graduate school, I'm a very biased sample, and I know lots of others -- very bright, talented, capable people -- who found it wasn't worth it for them. I don't think I would attempt to give advice even if I thought I could perfectly distinguish those who would find great personal success from graduate school from those who won't, and it's perfectly clear to me that I'm far from a perfect distinguisher.

Where possible, I try to give facts. Inevitably, people who find both work and graduate school compelling options want to know how difficult it would be to switch from working back to school. My take was that at the application level, a year or two working generally, at worst, does minimal harm to an application. Your professors still remember who you are well enough to write useful and informative letters, and your academic skills are assumed to have not gotten rusty. Coming back after an extended period, however, might make the application harder to judge.

The greater difficulty in switching is that the longer you work, the harder it can become. You get used to a real paycheck instead of a subsistence wage. Who wants to move again, uprooting their life (friends, relationships, etc.)? And you probably start to become attached to your job and your co-workers in various ways. [Interestingly, the same sorts of issues can arise for people who are thinking about academic jobs vs. research labs/other jobs after their PhD.]

Happily, the student seemed to not need my most important advice -- that both possibilities offered him great opportunities for success and happiness, so he should not stress about making a choice that was "wrong".

Does anyone have further, general advice for those facing this decision?


Anonymous said...

I went through a similar situation last winter. I was graduating in the spring, and was torn between graduate school and industry. My final choice: industry. My case may be unique, however.

I spent every summer of my undergraduate years doing research. I had 0 industry experience. The more I thought about the kind of research I wanted to do, the more I realized that my desire was to work on the vast bridge connecting theory and application. Finding theoretical solutions is interesting to me, but only if they would have a real-world impact. My problem: I had no idea what kinds of real-world problems people were dealing with. There's a difference between reading about the problems practitioners face in the real world (while you're in academia), and actually attempting to solve them in a day-to-day job.

I've been working for about 6 months now, and I am 100% confident that I made the right decision (for me). From dealing with system-level algorithms, to data storage problems, to high-level algorithms, to statistical analysis -- these real world issues let me view problems from a different angle. I still plan to go back to graduate school, but I'll be returning with a new perspective (and, one could argue, a more "productive" attitude).

Anonymous said...

Coming back after an extended period, however, might make the application harder to judge.

Is it really the case that actual, real-world work experience is harder to judge than whatever you have to go on for undergrad applicants (GRE + GPA + college/summer job experience)? IME, the similarity between undergrad school and grad school mostly ends with the word "school"; grad school is much more like real work, and thus it seems to me real work should be a better predictor of grad school performance than undergrad work.

I worked for two and a half years after undergrad. You're right that it's harder to get letters from undergrad profs, but my letters from my managers were glowing and I got into a good program.

Unknown said...

I can only speak from a perspective of someone who worked for ten years before going to grad school.

It sounds like your student could do whatever he wants to do, and if he wanted to go work and then return after a couple years, he wouldn't run into any barriers.

However: Like you said, you get used to a regular salary. Even a couple of years of this would make you accustomed to it, and it will be harder and harder to give up the longer you do it.

Without a graduate degree, the work you get will be interesting, but probably not challenging enough for a bright person. The one thing that might help (and will subsequently make it easier to return to grad school) is if there is research to be done in this position.

Take it from me: don't get content if what you really want is to be challenged. Whether he can internalize this at this point in his life is another question.

Anonymous said...

I worked for 2.5 years before heading back to grad school.
Here are the pros and cons (purely from personal experience. Extrapolate at your own risk!)

Pros (of Working before Grad School)
1) Better perspective on how the world works and how marketing, finance, strategy interact in any commercial enterprise.
2) It gives you more options to branch out to Law/MBA etc (most of which require a few years of experience). In other words, you will come back only if you really like engineering. Otherwise, you can move into an alternate, unexplored trajectory.
3)You learn to work with large systems with large existing code bases. Something, that you rarely do in academia. Realize that technical solutions can take you only so far. Thus I have a more respect for people who do "fluffy" work than my peers that came directly to grad school.
4) Money saved for 2 years has made my grad school a much more pleasant experience.

1) Getting back to theory courses is hard. My first semester, we were talking of convergence in a probability class (at week 4) and I was still struggling to get my hands moving quickly to compute expected values in a test.

2)I didn't move cities. But I realized that I had to downsize my life and spend far less to make stipends work. In effect, it meant meeting my old friends at restaurants/bars less regularly.

3) If you don't end up with a researchy position post grad school, the opportunity might seem excessive in hindsight. More excessive than, say, if you had gone to get a PhD directly.

Anonymous said...

I have started a job in an industry lab recently, and wanted to throw my perspective.

I took up PhD (in a top 10 school) just after finishing Masters in another university, so I can't really comment on going to an industry job and coming back (I did work for one year after bachelors, but that was a short period).

My original plan was to get Masters and join a regular job like everyone else, but the idea of doing cutting edge research working on tough problems lured into PhD. When I started, my only goal was to get a faculty position.

But, in the end, I chose an industry position. I didn't even apply for academic positions. A few reasons

1. The grind of tenure feels exhausting. After seeing many young profs working day and night to get their tenure, I didn't feel like going through that immediately after PhD. Also, the academic hiring scene right now is extremely tough.

I understand the benefits of tenure, and the flexibility of choosing many research ideas, but it just isn't for me right now.
2. I interned four times at research labs in the industry and the work is amazing. The satisfaction of putting your research stuff into real products is a great feeling that I really liked. You can still write papers, work with academia, travel etc. So, for me it looks like the best of both worlds.

Of course, industry labs always have the risk of company choosing to cut down research costs etc.

P.S. I am a regular reader of your blog, and really like it!

Anonymous said...

I had the unusual privilege of pursuing the doctoral portion of my graduate education and holding an industrial position simultaneously. Following this, I am now in a faculty position because I decided I want the freedom to do open-ended research.

Looking back, I find that the answer to the direct vs. indirect path is closely tied to whether one seeks an academic position at the end or not. Within academia, everything is measured in terms of years since PhD and the 'track record', which may ignore all other benefits of experience gained in industry beforehand. Moreover, in the American system, one may be going through the rigors of tenure track at an age when one's peers outside are onto the next stages of their lives and are 'settled'. Clearly, such considerations are highly personal but there is something to be said for the straight arrow path if one were intending to go through the standard academic track at the end of the PhD.

Anonymous said...

I am currently a PhD student in my 4th year. When I applied to grad schools, I also had several interesting and lucrative job offers.

The most salient advice I remember getting at the time was this: go to grad school only if you really love the subject. Otherwise, it may well not be worth the 5+ year commitment (either time-wise or money-wise, in terms of the relative pay difference).

(This shouldn't necessarily be the final word on such a decision, but no one else had mentioned it yet.)

Yuriy said...

In some disciplines within CS, going to work for a couple of years is actually a positive for applying to grad school (e.g., software engineering). On the other hand, while it may be assumed that your academic skills do not decline, it's amazing how hard it is to go back to test taking / note taking / problem setting after just 1 year away from school.

Anonymous said...

I am a faculty member who has served on the graduate admissions committee at my school for several years.

This is not an argument in favor or against going to industry or grad school, just an additional data point. If you go to industry and re-apply to graduate school later on, I think you face an uphill battle. If the work you did in industry was challenging, and gave you a perspective on what you want to accomplish in graduate school, you will have no trouble getting in. If you talk about how quickly you wrote web apps for your employer (and get letters of recommendation from your employer saying the same), and talk about how you want to go to grad school because you are bored at work, you will likely not get in to grad school.

Anonymous said...

I don't understand the comment about letters. I am pretty sure Harvard will hold onto reference letters for you, keeping them private from the applicant but available for applications. Ask for letters when the writers' memories are fresh. (I know that some schools ask for different information, but the letter itself is the hardest to put together.)

"Interestingly, the same sorts of issues can arise for people who are thinking about academic jobs vs. research labs/other jobs after their PhD."

Since there aren't any academic jobs any more, this dilemma fortunately no longer arises.

Seriously, if you are thinking about graduate school, think about where it is leading you in your career. Yes, the market should hopefully have recovered five years from now, but it will still be more difficult than it has been historically, because extended postdocs will have become entrenched in computer science.

Michael Mitzenmacher said...

Anon 2: I agree with Anon #9 about letters from managers, in the following senses. First, keep in mind that odds are the professor reading the letter from your manager doesn't know your manager. This makes the letter less valuable. If Prof X from Berkeley tells me, "This is the top student from my class of 50 this year," I know what that means. If your manager tells me you're the best of his 50 employees, that's much less informative.

Similarly your manager doesn't necessarily know what we're looking for in a graduate student. As Anon #9 points out, if we get a letter saying you write great web apps, we're not getting the information we need. If you don't have a research-oriented job, letters from managers often contain little information.

Anon #10 -- at most schools you can ask for a letter from a faculty member even if you're not applying anywhere for your folder. There are two issues. First, such letters tend to be more generic, since they're not for a specific purpose (getting into grad school at X), and hence often less informative and valuable. Second, on the receiving side, the more dated the letter, the less informative it is. Hearing what you were like two years ago gives some information, but not as much as a current letter.

Thanks to all who have given interesting and useful comments so far (and please add more!).

Anonymous said...

I went to grad school and the one I advice I have for everyone thinking of it is: read "PhD Comics". It is a pretty accurate description of the grad school life.

Luke said...

This is always a tough and very personal choice.

Some people could be happy in either place. For these undecided folks, I might nudge them towards industry as far as it gives them more options in 1-2 years. Many people change jobs or go back to grad school if they're not happy.

Some don't have the patience and focus for a multi-year thesis and should go to industry, where things are faster-paced. Some have a tough time fitting in with corporate culture and the more individually-structured time in academia is a better fit.

What do I wish I would've known as an undergrad weighing my options? I didn't realize how liberating a paycheck is. I had no idea corporate culture could be so positive (the stereotypes of office politics and bad managers don't hold at top companies). I didn't fully appreciate how nice a campus social life is.

Anonymous said...

My 2c.

Of the people I knew at grad school who had worked before joining grad school (at a top university), I think the "success rate" was noticeably lower. People who have worked often seem to have lower tolerance for research that doesn't seem immediately applicable. They have less patience and are less happy "wasting" their time working on something and not succeeding. They are also less used to attending classes, and don't necessarily remember some of the stuff they learnt as an undergrad. It may also just be that they have a clearer picture of what they want from grad school and get disappointed when the expectations are not met.

I would also second the opinion that grad school is a lot of work, and is probably not worth it unless you really enjoy what you study. In other words, if you are working hard at something you don't enjoy, you might as well be making a lot of money doing it, instead of getting a phd stamp.

Charles Sutton said...

I spent a year in industry before graduate school and (recently) getting a faculty job. I'm glad that I didn't go straight to graduate school, because:

(a) As others have noted, the paycheque is comfortable, so you do have be willing to break your routines and uproot yourself to make the switch. But the fact that I had to do that helped me to feel comfortable that I was doing grad school for the right reasons, rather than as a "default" choice.

(b) The programming skills, "teamwork" skills, and confidence that I gained in industry really did carry over to help me in my (moderately applied) research. Maybe the carry over would be more useful for some research areas than others.

(c) There were unique things about my industry position that did make it research-relevant.

That said, I agree with some of the cautionary responses on this thread as well (e.g., Anon @ 4:07 PM).

Anonymous said...

I think it is very important for professors to tell the student that going to grad school and doing a PhD does not mean that you will get an academic job and will be free to to the research of your choice. It is not only difficult to get an academic job but nowadays it is not even a possibility unless you are a female or some famous candidate from a top-10 school. No matter how good are your publications.. if your school is not in top-10 you will not get an interview... So IMO doing a Phd is worth only if its done from a top-10 CS schools.. period. If you go to any other school, chances of getting academic jobs are ZERO.

Stefan Savage said...

A few comments:

1) As a counterbalance to some previous comments, in my experience, students who have had some time in industry are frequently stronger graduate students that those who come right out of undergrad (and FWIW, I actively look for such candidates). They tend to be more mature and they've also given more thought as to _why_ they want to be in graduate school instead of defaulting into it. Indeed, one of our best recent students came to us after practicing law for several years. I think the biggest liability such students face is getting good advice on how to manage the application process (who should be writing letters for them, what they should be saying, etc)

2) Wrt making the grad school choice, I think there are three choices: MS, PhD, and industry. The MS degree is frequently thought to be a way to put your toe in the water for graduate school while improving your salary prospects in industry. It can be effective at both, but there are quite a few gotchas here. Among them, at many schools its much harder for MS students to get involved in research and at many (most?) schools it can be harder to go from the MS program into the Ph.D. program if you so desire. The Ph.D. program is a great choice iff you really like doing research. Its also a near-requirement for a research job which is one of the few career choices that can provide for high levels of independence. I think industry is best for students who get their value from direct impact on the world and/or who are economically motivated.

3) Finally, I think we should understand that the research job market is complex. One should not judge what the market will look like in 5-6 years by what it looks like now. When I entered grad school in 93/94 the academic job market was worse than now. When I went on the market things were booming. Its worth understanding that in addition to growth/shrinkage driven by undergraduate interest, there is a cyclic component driven by large demographic bubbles in the faculty population (our ages are not homogeneous :-). There's one bubble in the wake of PC era and another in the wake of the Internet. Each of these bubbles in turn will drive mass retirements. Also, to the last anonymous comment (only students from "top-10" schools get interviews) this is not supported by the data. It is indeed the case that some schools on this list have a high rate of academic placement relative to their graduating population, but many in fact do not (moreover these rates are highly skewed by area). Moreover, there are faculty candidates interviewed each year who are receiving their Ph.Ds from schools outside the "top-10".

Anonymous said...

"these rates are highly skewed by area".

How many men from non-top 10 programs working in TCS have gotten academic jobs recently?

Anonymous said...

^ Do you consider UT Austin top 10?

Geoff Knauth said...

I would absolutely recommend a good student go to graduate school. Companies come and go, but advanced degrees from institutions with a good history become more valuable as you get older. Think twenty years down the line. Would you rather be in your forties with or without graduate education under your belt? Twenty years goes by really quickly. I had a twenty year run of good luck with great jobs, then found out the hard way that jobs get much tougher to find in your forties, especially if you are geographically challenged. You can go back to school in your forties, but you might have a family and "life" might get in the way. Go to graduate school when you are unattached, young, and full of energy. It will only increase options available to you in later life. If you don't go to graduate school, then figure out a way to have forty years' of savings and retire at forty, because that could be closer to what you face than you think.

Another reason for going to graduate school is, if you surround yourself with smart people, some of it is bound to rub off, someone is bound to appreciate your own gifts, and throughout life you'll have smart friends. Companies are very hit or miss, and money is no substitute for family and friendship. That was a recurring theme at my 25th reunion in 2008.

Anonymous said...

Is it really true that female students have better chance in getting an academic jobs than their male counterparts ?

Does anybody have any statistics on female TCS students recruited recently ?

Anonymous said...

Is it really true that female students have better chance in getting an academic jobs than their male counterparts ?

No, this is actually a myth, perpetrated by male TCS students. In recent memory, I can only think of a handful of female TCS students who actually got academic/research lab jobs. In fact, a much higher fraction of my female TCS classmates dropped out of theory (started working for industry, changed fields, moved back to their home country, took a predominantly teaching job, etc.) than my male TCS classmates. And I went to a top 5 theory school, so it is not that being female and in a top 5 school automatically gets you jobs.

Claire said...

I worked for a year in industry before coming to graduate school, and I would highly encourage other students to consider doing the same.

Caveat: I'm generalizing heavily from a small set of data points, here, so take my claims with a grain of salt. :-)

Pros of a job include:
1) taking mental break from academia. I know several people who went straight to grad school in a number of disciplines and promptly burnt out.
2) making some money, which allows one to have a cushion for unexpected expenses.
3) learning adult habits/skills. At least as a Harvard undergrad, one tends to live in a dorm for four years and thus may not have much experience in operating a stove, budgeting, scheduling the cable guy, making one's own doctor's appointments, getting out of bed before 10 am. Knowing how to operate a stove is a pretty instrumental skill for surviving grad school, in my experience. This also applies to learning how to sit down and do work for N hours a day in an office.
4) motivating one's work. Admittedly, I do PL and software engineering, which means that working on a large compiler project with its bug database, requirements, regression test suite, backwards-compatibility requirements, etc really motivated my work now in improving software quality and the programmer experience.
5) knowing that one really wants to go to grad school (instead of just going because it seems like one should).

Cons of a job:
1) jobs can be boring.
2) inertia. The Real World gets pretty cozy, with a regular pay check and an apartment and a car and friends and hobbies and a city in which one is comfortable, a promotion, a raise...and it gets hard to leave. And that's not even considering people who start families (which I didn't, mercifully).

I don't think it's as dangerous to take a year or two off as some people seem to think it is - most of my good college friends took a few years off to work, travel, volunteer...and most of them are now back in school (none in CS, but in general), and are happier/more educated for it. It might be harder to apply if you've waited 10 years, but since I didn't, I can't really comment ... I didn't have too much trouble waiting a year at all.


PS Mitzenmacher - that blog post you link from your website about how scary you are in undegrad algorithms: that was me, several years ago, now. Hi!

Michael Mitzenmacher said...

Claire!!! Hi! First, thank you for your detailed and well-thought out comment on this post. I hope you'll have occasion to comment again on other issues. But also, thank you for writing that blog post about me years ago. As I'm sure you're aware (and in all seriousness, no sarcasm), it's one of my favorites of all time, and I'm keeping my copy of it. My kids will love it someday, though I may have to explain to them what a "Bond villain" is at that point.

I hope that, in retrospect, I'm somewhat less scary. And I'm always thrilled to hear from or about Harvard students continuing on in their careers, especially graduate school. We wish you great success!