Thursday, September 9, 2010

So, you want to go to grad school?


Every year I am approached by students asking about grad school in Computer Science. I generally sit down with them for an hour or so and go over all of the details of why you should go, what the tradeoffs are, where you should apply, what it takes to get in, and so forth. I figured it would be a good idea to write some of this advice up in a blog post so I can capture it in a more permanent form.

In this post, I will talk about why to do a PhD in Computer Science, and why not to do a PhD. Assuming you've already decided to go to grad school, I've blogged previously about how to get in. Later on I'll blog about where you should apply.

Masters vs. Ph.D.

First off, when I talk about "grad school," I mean doing a PhD. Many students ask me about doing a Master's degree after college. I don't generally recommend students from good CS programs do a Master's in CS, for several reasons: (1) it's expensive, (2) you can learn the same material as an undergrad, and (3) doing a Master's isn't useful for deciding if you want to do a PhD -- it is a totally different experience. M.S. programs generally require taking a lot of classes, so they are not at all like being a PhD student (where the focus is on research). PhD programs don't generally care whether you have a Master's when you apply; in fact, some schools seem to prefer taking students straight out of their undergrad degree.

The only cases I recommend doing a Masters are for students that aren't quite prepared to get into a top-ranked PhD program, for example, because their undergrad major is in something other than CS. (Note that if your undergrad major is in an area closely aligned with CS, such as engineering, math, or physics, or you took a lot of CS classes despite majoring in something else, you probably don't need a Master's.) A Master's can also benefit students coming from foreign universities. Doing a Master's at a good CS program in the US is a good way of getting a letter from a well-known CS professor to help you get into a PhD program.

Why do a PhD?

Of course, this is the most fundamental question. I'll try to articulate the pros and cons below. First, the pros:
  • Lots of freedom. PhD-level research is all about defining a problem, solving it, and convincing everybody that your solution is a good one. Half of the challenge of doing a PhD is deciding what problem to work on. It is really about carving out your own niche in the field.
  • Working for yourself. Once you have a PhD -- and even during the process of getting one -- you are able to be your own boss. Rather than working on someone else's vision, you are the one to define the vision. This is especially true if you pursue an academic career after grad school, but is also the case in many industrial research labs. Typically, people with Bachelor's and Master's degrees aren't afforded so much freedom. 
  • Working on the hardest problems. PhD research is about opening up new avenues of enquiry, and working on problems that the rest of the world hasn't even articulated yet. If you do it right, you can have tremendous impact.

Why not do a PhD?

Of course, doing a PhD is not for everybody. I have seen quite a few students enter a PhD program, spin their wheels for years on end, and leave without finishing their degrees or doing much of anything. I've even see people get a PhD without making a mark on the academic community, just barely doing enough to get a thesis signed off by three professors (this is easier than it sounds). These people shouldn't have done a PhD at all -- they would have been better off going straight to industry, making a lot more money, and probably being much happier in their jobs.

The only reason to do a PhD is because you love doing research. If you don't love research, don't bother -- it is not worth the time, money (in terms of opportunity cost vs. making a real salary in industry), or stress. Doing a PhD is stressful, if you are doing it right -- you are in constant competition with other academics to publish your results in the top venues, to make a name for yourself, to get recognized. If you harbor ideas of lazy days sitting in the coffee shop pondering the universe, you are dead wrong. (You can always approach a PhD this way, but you will probably not be very successful.)

"But," you say, "I don't know if I love research -- I've never done any!" Then why are you considering doing a PhD at all? The only way to find out is by doing research, preferably as an undergrad. If you screwed up and graduated before doing research, try to find a research assistant job in a professor's lab, or do a Master's (see above). Be warned that most Master's programs are very course-intensive, so you will need to work extra hard to do some research on top of the courseload.

Another downside to the PhD is that is it extremely unstructured. This can drive some people crazy. The nature of research is that it is open-ended, and there are often no clear guideposts as to what you should be working on each day. Also, your PhD advisor may or may not mesh with your personality -- they might be too hands-off, too hands-on, out to lunch, too stressed about getting tenure, etc. Your experience in grad school will depend a lot on how well you get along with your advisor. (Let me take this opportunity to apologize to all of my current and former students for what they have to put up with.)

Doing a PhD can take a long time. Nobody finishes in four years. The typical time to completion is around five or six years, but there is a long tail -- I reserve the term "paleo-student" for someone who has been at it more than 10 years. See the Taulbee Survey for some data. The time to finish your degree can be taxing, since all of your friends have already gone ahead and gotten married, had kids, bought a house, etc. while you're still living in squalor with four roommates who haven't bathed in a week. Eventually your parents and loved ones start wondering what the hell you are doing with your life. The brilliant comic strip Piled Higher and Deeper uses this as a recurring theme. My advisor used to say that "doing a PhD costs you a house," which is just about right if you consider the amount of money you could have made being in industry for the same amount of time.

So, should you do a PhD, or not? If you think you are up for it, you can always try it for a couple of years, and if you dislike it, go get a job in industry instead. Unfortunately, it doesn't quite work the other way -- moving from industry to grad school is much harder.

Taking a year off

A lot of students tell me that they plan to get their bachelor's degree, work in industry "for a year or two" and then apply to grad school "later." If you are serious about going to grad school, I do not recommend this approach. In my experience, it is quite rare to make the jump from industry to grad school. First off, industry pays so much better than the PhD student stipend that it is quite hard to make this transition. Also, to get into a top PhD program, you need good letters from CS professors, and letters from industry don't really count. After you've been gone for a couple of years it's hard to get those stellar letters from the professors that may have loved you back when you were in college; newer, brighter, more energetic students have taken your place and you are long forgotten (although maybe Facebook will change all that). Industry experience rarely helps a graduate application, especially if you're some low-level engineer at a big company writing tests all day.

That said, taking time off after college can be a great experience. I took a year off doing research at different universities (University of Cambridge, University of Glasgow, and Vrije Universiteit in Amsterdam) after finishing college but before applying to grad school. It was a great experience and it bolstered my grad school applications since I stayed within the academic sphere.

Another approach is to get into grad school and then defer admission for a year. Most schools will let you do this (although they may grumble a little, or even make you re-apply, although this is usually a formality).


46 comments:

  1. Thank you for this interesting post.

    "Nobody finishes in four years". Most countries here in Europe (particulary post-Bologna) pushes to have 3 year Phd programms (after a 5 year graduate degree). How does this reflect on the quality of the final thesis? Do you consider these PhD inferior, or not-so-mature?

    ReplyDelete
  2. Ah yes. I should have pointed out that this article only applies to PhD programs in the US (in Computer Science). Depending on the country, PhDs from Europe are vastly different in terms of content, structure, and timing. A UK PhD, for example, is rarely seen as equivalent to a US-based PhD, precisely because it is so short and does not typically involve coursework.

    ReplyDelete
  3. Matt, I am a grad student myself and I want to reemphasize that you can only do a good PhD (where you actually develop as a researcher) if you are excited by your field of research.
    1st/2nd year is fine, but it is very difficult to remain focused, if you are not psyched by your research.

    ReplyDelete
  4. My understanding is that in the UK for example you have exactly 3 years to finish your PhD or the funding will be cut off.

    This is basically like a good US Masters thesis.

    Personally, I did 33% of my eventual PhD thesis research in the first 3 years and 67% in another 3 years. This "backloading" seems typical (partly due to taking classes and gaining experience + the pressure to graduate). In any case 3 years seems to short for MOST students (there are always exceptions).

    ReplyDelete
  5. I think taking a year off is actually a great way to make sure you really are interested in doing research, and totally reasonable if you go to a company which does some research or works on problems that are of interest to academics, like Google or Facebook. There are a lot of people at these companies that still have ties to academia, and can write respected recommendation letters.

    People who don't even know what the real world is like can't make an informed decision.

    ReplyDelete
  6. Anon #1 - I agree that being excited by your area of research is critical. But keep in mind that ALL grad students (nearly without exception) go through a period of really, totally hating what they do. (Piled Higher and Deeper would not exist if this were not the case :-) Sometimes this slump can last for a year or more, and I've seen more than a few grad students quit as a result. Research is not always fun, or exciting, or sexy - usually there is this critical point about 3-4 years in when you are done with your first couple of projects and have to decide what's next, and hit a brick wall. The key is persistence.

    Anon #2 - UK PhDs do not generally make one competitive for a faculty position in the US, without some additional postdoc time. Note that a postdoc is increasingly common for US-based PhDs as well, in part due to the bad economy and lack of academic jobs.

    Neha - Being at Google now myself I agree COMPLETELY that real world experience is super important. The problem is that (for whatever reason) there seems to be a huge barrier to people moving from industry to a PhD program. This is true even at more "academic" places. I've never seen one of my students who goes off to industry "for a year or two" end up going to do a PhD - probably because they are having enough fun and making enough money in the real world. So I still strongly encourage students who want to do a PhD to avoid the temptation and stay in the world of make-believe until they finish grad school :-)

    ReplyDelete
  7. It's unfair to characterize a UK PhD as "a good US Master's thesis". I am an American who has been through a US PhD and I am on the research staff at a UK University. The PhDs at my institute are awarded only for serious original research. The differences are more subtle than you think:

    -Most UK students come in with a one- or two-year master's degree -- which in many cases entails a significant research component. Furthermore, although the advertised time to completion is three years, there is usually no problem obtaining extensions or funding for another year. It's usually only in the fifth year that institutional pressures seriously kick in. By contrast, US PhD programs admit students with only a bachelor's degree and require coursework for the first year or two (often leading to a Master's on the way). When you compare programs in this way (UK 1+3+ to US 5+) the time difference is not all that great.

    -The *main* difference is that UK PhDs are, I think, somewhat less likely to work on side projects or internships (although this is not unheard of). They work straight through on their thesis topic for three years. It is true that in most cases a student who finishes in three years on the nose may have had significant direction from their advisor -- i.e. they may less experience finding their own problems (although those who take 4-5 years are probably just as good at this as a US student). So it might be fair to say they are slightly less well-rounded. But generally speaking that is solved by a year or two at a postdoc. (As a side note, this is more humane than the US system in a certain way: students cannot linger for 10+ years, so cannot become paleo-students. Those who finish in 3 years and decide they hate academia get their lives back, and those who want to stay can take a bit longer and receive additional training at a postdoc. Win-win.)

    -I agree that Americans seem to regard UK PhD's with suspicion, but frankly I have come to believe that this is due to provincialism as much as anything else. A colleague of mine went directly from a UK PhD to American faculty (yes, this is rare; he is American) and is doing extremely well -- interesting research with lots of grants, students, and publications.

    ReplyDelete
  8. Curious that the US should have such a low opinion of UK PhDs. However if you (Matt) hold this view it seems reasonable - you've been to two UK universities...

    The one thing I will note is that if you are deciding whether to do a PhD think very hard about what you enjoy and are good at. If you don't like (or struggle with) large (where large is at _least_ 7500 words a week) amounts of reading and writing in your subject I would say it is not for you. In the UK you can do research oriented masters and effectively a PhD feels a lot like taking a project you would do in one of those and then stretching it out for 3 years. If you didn't love and do well at the project parts of a BSc/MSc it is likely you won't love and do well at a PhD either. As everyone points out, your relationship with your supervisor can make a huge difference while you are doing it, as can the life you lead outside your PhD and what you want to do after you get your PhD. If you can't see yourself going into academia a PhD is probably not for you (this is not to say you HAVE to go to into academia afterwards - most graduates don't but if you should at least see yourself being able to do that when you start).

    ReplyDelete
  9. Thanks for the very interesting post.

    It would be nice, given your experience, if you could share your thoughts also on another topic, i.e. the after Ph.D. (post-Doc/Assistant Professorship), addressing questions like:
    * better a post-Doc in a good/top Lab/University or an Assistant Professor position in a minor/small Lab/University?
    * how to get a post-Doc/Assistant Professor position? What is most valuable to get it?

    Thanks!

    ReplyDelete
  10. Hi Matt, I have wondering myself about this all the time. I did a master. I was luck and it was not just a "classes master", but thanks to my professor who pushed on research direction, it was much more fun: publishing, do a research exchange period, organizing seminars and summer schools, etc.. After that I have been working at a research lab [in fact I'm in France now, traveling and working in other Universities :-)].

    So, I'm pretty sure that I can do fine in a PhD program [and a lot reading this blog are...], but the question still remains: do it or not?. The problem I see have nothing related with actually "doing it" [which, as Matt Might states, is not that hard http://matt.might.net/articles/successful-phd-students/ or http://matt.might.net/articles/ways-to-fail-a-phd/], but it's to think in "future". Not all people wants to keep doing research as a profesor and PhD seems pretty much just useful in Universities [or, in other way: in my work I do pretty much the same than people with PhD, except that I can't apply for government grants because I don't have the PhD]

    Finally, I said myself, I still can create my own company... but you know, that you have invested 4 years of living as student in the PhD, that you are no sure you want to keep other time on that... (or, to loose another house)...

    Would be nice if you can talk about life and opportunities after the PhD...

    cheers,
    Victor

    ReplyDelete
  11. I've heard the line about avoiding the temptation of the real world from a few people, and I don't really agree with it.

    A lot of grad students are kind of hazy about why they're doing what they're doing, or what exactly it is they're supposed to be doing. The people I've known who've come back to academia 5-10 years later have had a strong reason for it, are used to longer-term self-motivated and collaborative projects, and have some perspective on real problems. I wouldn't be surprised if those with significant pre-PhD work experience have a much higher academic job placement rate vs. those who go straight through from undergrad.

    Personally, I found that working with a startup in the later part of my Ph.D. program helped inform my perspective: how to choose a problem that can have real impact as opposed to something artificial that seems difficult hence impressive; and why I might want to work more on research problems as opposed to the typical development grind. But it would have been a lot more useful to have had that background coming into the program.

    ReplyDelete
  12. Hi Matt, and thanks for another insightful blog post.

    I think I agree on most terms but would disagree on viewing this as generalized advice.

    For one, I think industry experience before you enter a Ph.D. program can be very beneficial, in terms of focus and discipline, but also in terms of choosing problems.

    For those with less drive, the choice of an advisor becomes more important. The pre-structuring of work and guidance you can receive can vary widely by advisor, so not all advisor/advisee pairings are born equal.

    Some notes on comparing a U.S. Ph.D. with a German Ph.D. here: http://osr.cs.fau.de/2010/06/08/the-german-ph-d-system-explained-in-one-post/

    Cheers,
    Dirk

    ReplyDelete
  13. Thank you so much for this interesting post. I could get the insights about doing a PhD.

    My goal was/is to create something useful in computer science. But I screwed up my high school years and went to a bad college (but still got into my fav course - computer science). Some how studied the subjects which are interesting to me with absolutely zero guidance from college and I am now working as a rails programmer in India. just passed out from college 3 months ago and I want to do PhD from a very good university. Since my undergrad education in computer science was bad , I prefer doing a masters course to strengthen my fundamentals before I jump into PhD. MTech Computer science would be my course and I will get them in a very good college here in India (IIT or IISc). And then I prefer doing PhD in USA.

    Do you see anything wrong in what I am doing. I love to read and think on a lot of problems and this feeling convinces me that I am the right one for PhD. Please comment on this.

    Again, thank you so much for writing this post. Bookmarking it now.

    ReplyDelete
  14. Damn, now I wish I paid more attention in college.

    ReplyDelete
  15. Ok, lots of good comments here...

    Anon re: UK phd.... It may be unfair to characterize a UK PhD in this way but the fact is that this seems to be the dominant perception in the US. The two systems are difficult to compare.

    Anon re: postdoc vs. Assistant prof... Great question. Worth its own post. I'll do one on that soon, but the short answer is that it depends a LOT on the circumstances!

    Victor... The key question is whether not having a PhD is holding you back on the career path you want to be on. Sounds like you are quite academic/research focused so I think a PhD is extremely valuable in that domain. Otherwise you will always be working on someone else's research rather than your own. This is not to say you need a PhD to have a happy and successful career... :-)

    Sameer, I agree 100% with you and in fact really like to see PhD applicants with *appropriate* industry experience. One of my new students is coming to Harvard after spending a few years doing a startup in the wireless space, work very relevant to my interests. Working on the menu bar UI in Microsoft Excel is not that helpful. My observation is that undergrads who SAY they want to take a few years in industry first rarely make it to grad school of their own volition. Not that industry experience is bad.

    Lakshman.... If you want to do a PhD in a good school in the US, then do a Masters in the US. Otherwise IIT is well recognized in the US, but you will be in competition with many more applicants from those schools. A good US masters (say from Stanford or CMU) should be easier to get into than IIT!

    ReplyDelete
  16. Masters from Stanford or CMU is really tough to get - at least thats what people say. Schools like stanford or CMU asks for good UG record/grades/paper-submissions/extra-curriculars/good-UG-college etc., (which in my case is really bad) .. am I right ?

    Or am i getting Mislead by others comments...? Please guide me

    ReplyDelete
  17. Lakshman, masters from US will cost you ~50K USD.
    And Matt is correct. Getting a seat in a good master's program is difficult than getting into a masters program in US.
    The only criteria for MS in USA is above average grades.
    If you can afford the huge tuition in USA then I would say, go for it.

    ReplyDelete
  18. Hi,
    it was interesting to read about the US and UK phd system. What would you say about the german "Doktorarbeit" in comparison? To me- beeing used to the german system- especially the possibilty to do a phd without a master first was surprising.

    ReplyDelete
  19. You are killing my dreams though thank you for being so clear. You confirmed what I suspected. I'm in the "no research experience and currently working" boat. I knew I was going to have trouble getting in anywhere really great. I've been trying to keep my schools realistic or so I think. Right now I'm looking at Iowa, Misso, and UGA. I was hoping not to take a step back from where I graduated, NC State, where I got my BS and MS.

    You're right about getting letters from professors. I wasn't very friendly with my teachers during my BS or MS. Since I've been working for two years now, I definitely don't think they would give me a recommendation nor would I feel comfortable asking them. I was planning on getting letters from people I worked with, which I know is weak, but it's all I can do :(

    ReplyDelete
  20. Going to the real world before getting a Ph.D. is something that is definitely rare, but it's certainly something I recommend. I worked for over 4 years before going to Cal for my Ph.D. I believe that the perspective and experience was great for me. More than anything else, by the time I joined the Ph.D. program I was 100% sure that it's what I wanted to do and as such had no self-doubts through the process - something that plagues plenty of students.

    Matt's right that the transition is hard to make and most people end up not doing it. This is why many people tried to dissuade me from going off to work before my Ph.D. - my perspective, however, is if you really cannot tear yourself away from industry to do a Ph.D., then a Ph.D. is not worth doing for you.

    As for the anon who feels he has lost his opportunity to go to grad school because he did not cultivate a good relationship with professors as an undergrad. My suggestion is find an exciting startup that has spun out of academia where the professor is still highly involved on a day-to-day basis (e.g., Eucalyptus) and try and get a job there. Be upfront that your goal is to spend two years, impress the professor and get a great recommendation. Then work your ass off and actually impress the professor, spend extra time working on writing up your and your colleagues work for publication along with the professor. Even if you can only end up getting an "industry" paper as opposed to a real research paper, the experience will be very valuable and very helpful for getting into a grad school.

    I know of at least 2 great guys who used this approach to leverage themselves into top grad schools.

    ReplyDelete
  21. thank you for this article.

    In Russia, PhD de jure takes ~6 years (2 MS+ 3 PhD + 6-12 month of defence related stuff). But de facto it takes something like 7-8 years after your BS, but these extra years , you are not a grad student, you are "undegraduated" researcher...

    ReplyDelete
  22. Can you comment on valid reasons for getting a master's degree? I'm an industrial programmer whose academic background was math (completely self-taught in CS) and am thinking about using a CS master's program as an opportunity to develop depth in a specialty. I have no particular specialty now, and I know I can't expect much in the way of challenging technical work without developing depth in a particular area. (Career-wise, I've pretty much topped out as a generalist programmer at Senior Architect, and I don't like the work: too much pushing paper and telling other programmers what not to do.) Is a master's program a good way to develop depth in something like statistical analysis and big data?

    ReplyDelete
  23. I'm not doing a PHd, because I don't need an authority to certify that I'm innovative nor do I need anyone to provide me with the "golden" opportunity to pursue research. The credibility of such an authority itself is debatable. I believe in myself and that's my strength. After reading a few papers published, I have a feeling that Phds are treading the same path as MBAs with their sophisticated jargon.

    ReplyDelete
  24. Where are these magical Phd programs that don't require course work? Every single program that I looked at basically *requires* people to do the equivalent of a master's degree on the way to getting their Phd, including writing a Master's Thesis.

    The unstructured environment isn't the problem with going to grad school. The problem is that precisely ZERO percent of the professors and staff give a shit about the students. I went to an R1 school and, i swear I am not making this up, the hallway to the professors' offices was locked and not accessible to students, and administrators actually got pissed if you came to them with problems.

    The standard refrain was: "Go away. I'm too busy running a University to worry about the students."

    ReplyDelete
  25. to Anon ^above,

    WOW! You have abit strange experience. In Russia you have to pass through Master program before becoming a Doctoral Student.

    But ususaly your professors are really supportive, expecialy your research advisor.

    ReplyDelete
  26. HI Matt,
    great article, I do agree also that UK PhD is usually of lower level of overall knowledge in the field due to having to do it in such short time, sometimes after a 3 year Bachelor's program.. this is why in Europe we always have to do postdocs if we want to get into good faculty posts, I did mine in four years but i did 2 internships. anyway even in UK most PhD programs are now pushing for 4-years funding

    ReplyDelete
  27. Hi

    About do not need MS for doing a PHD? That is very unusual. Can you give examples? I have researched online about the admission requirements for most of the respectable Univercities and none of them state that a MS is not required. I even emailed most of them to make sure.

    Here in South Africa:
    In South Africa, an honours degree is usually a postgraduate degree in the same area as the undergraduate major, and requires an additional year of study after your BSc degree.
    After your Honours you have to do a MS degree which is additional 2 years and then only you will get admission to the PHD program which is additional 2-6 years.

    Minimum year of study breakdown:
    ================================
    Bachelors’ Degrees (from 3 years to 6 years of study, depending on course)
    Honor’s Degrees (1 further year of undergraduate study, requiring a thesis)
    Master’s Degree (2 years of post-graduate study)
    Doctorate (variable in duration with a minimum of 2 years, following a Master’s)

    ReplyDelete
  28. Anon re: "valid reasons for getting a master's degree" - I think you answered your own question. You are self-taught in CS and have a math background. In your case a Masters in CS *might* be a useful thing to do, to bone up your CS background, but ask yourself what your goals are. One reason might be to be a stepping stone to get into a good PhD program. If on the other hand you are doing it to get a better job in industry, I doubt this will be in your favor (financially or time-wise). Though there are some really idiotic companies that seem to think that people with an MS should automatically get paid more than someone with a BS, regardless of experience. Real world experience generally trumps a Masters.

    Anon re: "ZERO percent of the professors and staff give a shit about the students." This depends enormously on the school and I don't agree with your generalization. One of the reasons I like being at a smaller department, like Harvard, is that we have a lot of interaction with the students and even undergrads get heavily involved in research. Not all schools are like this of course.

    Anon re: "do not need MS for doing a PHD?" In the US, PhD programs typically do NOT require a Master's degree for admission. However, some schools technically admit students with only a Bachelor's degree into a "joint MS/PhD" program in which you first go for your Masters and then can stay on for your PhD as long as you are doing well enough. At Berkeley, for example, you were not formally considered to be a "PhD candidate" until you got your Masters sometime in the first couple of years of the program. But even without a Masters' coming in, you are still considered a "PhD student."

    This varies a lot by country. I suspect that South Africa is more like the UK in that regard.

    ReplyDelete
  29. reply on : "do not need MS for doing a PHD?"
    ===========================================
    If this is the case that a student get admission to PHD without a MS but actually the program is: a joint MS+PHD program and only after completing the MS be a PHD student then => it is in fact still a requirement to have a MS for a PHD at that partucular school. If the school let the student study a longer PHD program over more years to incorporate the MS as part of the PHD then in fact it is still the same as doing a MS first and then over a shorter time frame doing a PHD which then are the same at the end of the day.

    In my opinion:The quality of research produced after completing a MS or PHD are not linked to the amount of years spend in college but to have a researcher's mindset on using your creativity to find solutions to problems based upon theortical and practical knowledge and skill.

    ReplyDelete
  30. thanx Matt for the amazing post. It helped a lot.

    ReplyDelete
  31. I'm not doing a PHd, because I don't need an authority to certify that I'm innovative nor do I need anyone to provide me with the "golden" opportunity to pursue research. The credibility of such an authority itself is debatable. I believe in myself and that's my strength. After reading a few papers published, I have a feeling that Phds are treading the same path as MBAs with their sophisticated jargon.
    ============================================

    Hey there,

    I agree and disagree, like so many things in life :) I'll tease apart what's in my mind, perhaps that'll help a little.

    I believe, to most people, a successful PhD is about paper publishing, and the kind of job after graduation. It's a lot about what others think about you: how well you brand your work, how many people getting to use what you created, etc. For getting a professor job right after graduation, it can be about how hard your advisor banged on the table to get you accepted in that university. For getting papers published, it can be about how well-accepted you are in the community (double-blind is mostly a myth). To get tenure, it's about how much your fellow colleagues in your university like you. After all, they're not about to have to look at someone for the rest of their lives if they don't like you ;) But yet, when you walk down the street and you ask someone if they know about this and that famous professor, they probably don't. It's mostly a lot about a small group of people who spend time patting each others' backs, and hence it's a little sad actually. So, with respect to this political aspect of a PhD, I agree with you that it just sucks that it plays such a significant role :)

    But there's a different aspect to a PhD. It's about philosophy. I heard on NPR today, a guy who went to interview a hundred (!) philosophers and another (a professor) were arguing, with the professor basically saying that philosophers don't say anything useful. And the reply from the interviewer stuck in my mind: he said that philosophy is about clear thinking: it's about structuring one's thoughts, forming one's arguments about a particular situation. And the very thing that the professor is trying to do is, in fact, philosophy, and that's useful. And a PhD is essentially just that: clear, structured thinking. To me, it's about the training to think properly. It's about learning to spot the important ideas so that when they show up, you don't let them go no matter how much insecure people around you say how much they suck :) It's about what you want to do with yourself, not how others impact you to the degree that conferences and tenure-committees have today. That, I believe, is the true spirit of a PhD.

    ReplyDelete
  32. if family position forces me not to do grad studies - what can i do. I am now working as a software programmer and I need to keep up and contribute to computer science research community.. is this possible without going to college ? .. please advise me on this..

    ReplyDelete
  33. "PhD is essentially just that: clear, structured thinking." Wow. I could not have said it better myself :-) Nice comment.

    Lakshmanan - why would you want to do a PhD if you can't support yourself or your family while doing it? There are more important things in life than having a PhD. Being happy, productive, having a family, all of those things certainly dominate. You don't need the three letters after your name for any of those things.

    ReplyDelete
  34. But i want and would love to do research. the degree doesn't matter - i want to do research. how can i keep up and contribute to computer science research community

    ReplyDelete
  35. Anybody can do research! There is no magical blessing that you get from a University to allow you to do that. Being at a University helps, though, in terms of staying abreast of the field and working with other academics. If you are wanting to do research independently, you can...
    (1) Go to scientific conferences and workshops (potentially expensive, though)
    (2) Read scientific papers - most conferences post proceedings online these days, and authors often post papers on their websites anyway
    (3) Participate in research-related discussions via various email lists (depending on your topic of interest)
    (4) Volunteer at a nearby University to work in a lab or with a prof on a project for free.

    ReplyDelete
  36. Its true that the stipend for a PhD student is much less than the average salary of a CS engineer in the industry. But when one gets the doctorate degree, I wonder how much does the industry has in offer for him? Also, I would like to know does PhD means being restricted to academic sphere? Or, doctorates can go to the industry if they wish?

    ReplyDelete
  37. Muthiah - the Taulbee Survey (referenced in the post) has details on faculty salaries. Salaries in industrial positions tend to be somewhat higher (sometimes much higher). In general, the cost to doing a PhD is something like $500,000 if you consider the difference between going to industry vs. grad student stipend. Longer-term earning potential *with* a PhD is no doubt higher, but it will take a long time to recoup that cost.

    Having a PhD in *no* way restricts you to academic positions! (This may be true in some fields, like english lit, not in Computer Science or engineering.) Most CS PhDs go to industry positions in places like Microsoft Research, Google, etc.

    ReplyDelete
  38. This comment has been removed by the author.

    ReplyDelete
  39. Prof. Welsh, really enjoy your blog and discussion, I have been doing lot of soul-searching to go back to school and this covered all the reasonable points. I wish had this insight when finished undergrad 5+ years ago, since it was from smaller state school in CA and screw up that graduated with great GPA but not much research or research experience. Thought will work in the industry and go back to school.
    (well just like you mentioned it's always hard and it keep getting hard...)
    One thing that have changed in the last ten years is now the state schools are getting superb new professors in EECS (from Ivy league/top tier schools and very reputed research departments. Also it's interesting just like most other fields new CS faculty have usually done post-doc.

    ReplyDelete
  40. Thanks Matt, your look at both sides of the "world" certainly helped me a lot in deciding whether or not i like to do post-grad. Thanks for this thoughtful post, appreciate your time on it.

    ReplyDelete
  41. Hi Matt. I enjoyed this post. One of the first questions I ask every prospective PhD student is *why* they want a PhD. A lot of times I hear the response, "Because I love teaching, and want to teach at the university level." I'm wondering about your thoughts on that as a motivation for getting the PhD. Is this particular motivation a potential red flag, assuming the reality is: (1) Most people coming out with PhDs will not end up in academia (even if this is what they want to do) because their research record won't be strong enough to land an academic job, and (2) The metrics of success in graduate school will inherently reward those who love research, and not necessarily those who love teaching. Anyway, curious on your thoughts.

    ReplyDelete
  42. Hi Kim! You make a great point and this is something that I completely overlooked in my post. Getting a PhD is absolutely a prerequisite for being a university faculty member, even at teaching colleges. It's a sad fact that to get a *teaching* position you often have to be a stellar *researcher* first -- as we all know these two skills are not necessarily commensurate :-) So yes, getting a PhD in order to teach is a great motivation, though it's too bad that doing well in grad school often has little to do with teaching.

    ReplyDelete
  43. Hi Matt, I am a master's student in Sweden. I always think of doing my PhD after grad and I look forward to it. Here in Sweden, a PhD takes 5 years and they also need a master's degree as a pre-requisite. This is quite different from a 3 years PhD programs in other European Universities like the UK. It involves some academic work at the faculty like 20% teaching and there are also course works that one should do. There is also a half PhD thing called Licenciate that a student decides if he should continue to pursue his research. What do you think about this program relative to the UK and US ones?

    ReplyDelete
  44. nigatu:

    A PhD in Sweden is 4 years full time, but typically includes 20% institution work (TA, administration, etc) so 5 years. A masters' degree is usually requested, but not formally required.

    The thesis can be written in two styles; as a monography (start to finish), or collection of papers. For CS theses, the thesis are usually the collection kind, with an introduction of about 50 pages and 8-10 published papers from peer-reviewed conferences and journals.

    From a quality perspective I find the "collection" thesis interesting, since many different people have reviewed your research and it has been published in proceedings, compared to a monography that may be reviewed "only" by the opponent and thesis committe.

    Myself, I published papers and am currently rebasing into a monography to get a better reading flow.

    ReplyDelete
  45. Hi Matt,
    I am a junior right now taking a double major of Math&CS.I have done research one Summer ,in astrophysics.Are my chances at a top CS phd school so slim i should not even try?Because that would mean that i take a math phd instead.

    ReplyDelete
  46. I'm offer up these comments as a counter to a few things Matt had to say.

    I'm really happy I spent some time in industry before going to grad school: 1) I had money in the bank while a grad student which I could tap if needed (but I didn't) and it could grow through the power of investing and 2) I understood what real world problems were out there (worked for one of the hot systems companies of the 1990s) and I had perspective.

    To overcome Matt's suggestion that students taking a break from academia to work have a hard time getting good recommendations from academics, I asked profs toward the end of my tenure in undergraduate if they would write letters for me and submit them to service my undergraduate institution provided to hold letters of recommendation. They were willing to oblige.

    I also went to a non-tracked graduate institution since I didn't know if I wanted a MS or PhD. (I had done CS research as an undergraduate, but still didn't know if this was what I wanted to do. I knew if research was my route I would want to be in an industrial lab and not academia.) This institution is consistently seen as top ten CS school.

    In the end, I left after receiving my MS and doing research and passing my PhD exams. I was tired of my research and had a working relationship with my advisers that became very apparently broken about eight months into the working relationship and never was repaired. (Bar none, worst experience of my life thus far.)

    Reflecting on my experience, I am happy I got my MS for the following reasons:

    1) I learned a lot more because of it.

    2) I also feel I am a better job candidate as "MS or PhD preferred" shows up more and more in the job postings I see from companies I believe would be a good fit for me.

    3) I know where to look for new ideas and how to critique ideas given the history of work in specific sub areas of CS. So, I can act as a bridge between academia and industry, which many of my colleagues cannot do because they haven't been exposed to CS research in a significant way. (Paxos and the CAP theorem are two ideas that have come up in conversations over the past two years.)

    The research experience also had some good elements as it taught me a bit about presenting my work and methodology.

    ReplyDelete

Startup Life: Three Months In

I've posted a story to Medium on what it's been like to work at a startup, after years at Google. Check it out here.