Thinking of dropping out of PhD program, need advice

[TomServo]

Grad school is *not* what I was hoping it would be. I wanted to study GR and was fortunate enough to get accepted in a program with an advisor who is a gravity theorist. I have the best fellowship awarded to grad students at my school. I passed all my prelims on my first try. My grades aren't a problem. But I spent three miserable years grinding through courses recovering what I covered in undergrad (which I would have liked if I'd actually learned anything but I didn't), taking electives I detested because they were the only options that year (small program), teaching labs that take up a lot more of my time than you'd think (partly because of how they are scheduled), and didn't even get to take GR until the end of my third year. I'm starting my fourth year and don't even have a thesis topic (there are vague ideas of what sort of things I might research but they are vague).

The only things I've actually enjoyed during my time here are taking GR and helping somebody with some programming over the Summer (my undergrad was half-CS). Every year for the past several years "I'm just about to start research" but nothing happens. It's partly my fault for feeling burned out but also, like I said, I couldn't even take a class in GR, the thing I came to grad school to study, until the past Spring semester.

My personal life has turned to absolute crap, I've been seeing a psychiatrist and a counselor off and on since last Spring. I'm aware that my current misery is *partly* to do with my personal life and not grad school per se but these things in my life would *not* have happened if I hadn't come to grad school in the first place.

IF YOU SKIPPED ALL OF THAT, START READING HERE: So I'm considering dropping out of grad school, I have a background in CS, and I'm not sure how to weigh my options. I thought of maybe applying for some jobs or reaching out to recruiters and seeing what happens, but I don't know if this is a good idea. I was reminded this Summer how much I love programming, and would love to get a job in software engineering, especially something that uses physics. I still love physics and still want to study GR as a hobby if I drop out.

Has anybody dropped out of a PhD program (when they were in good standing in the program) and had success along these lines?

 

[Locrian]

Get out with a masters and build a portfolio of software projects on the way.

I would encourage you to explore things beyond just physics in software. Most coding jobs aren't very interesting IMO. But there are lots of coding jobs, and therefore the small percentage of interesting ones amount to a pretty sizable number. Restricting yourself to physics would make it even less likely to get one.

Best of luck

 

[Choppy]

Sorry to hear your experience isn't what you expected it to be. And based on your description, it sounds like you have a right to be concerned. You're going into the fourth year of a PhD program and you haven't done any research? Is your supervisory committee okay with this? Not that I have any experience in the GR realm of things, but in my experience, students generally get some leeway over the first year to take courses, investigate potential research topics and advisors, and pass the qualifying examination. Sometimes if the student fails the qualifying exam up to an additional year is allowed without much research progress expected (but even then the student is expected to have a topic selected and some preliminary background work done), and it doesn't sound like this is an issue for you anyway. Something just doesn't seem right about this, but it is what it is, I suppose.

Maybe there's more to the story, but if you can I would start by discussing these feelings with your supervisor, supervisory committee, graduate advisor, and/or the department's graduate program coordinator. You don't have to tell them you're considering quitting the program, but if you make them aware of your concerns, it may help to keep the next student from falling into the same experience.

It might also be important to try to avoid any sunken cost effect here. If you were to get on track with an awesome project in the coming months, would that change your mind? Things probably won't change unless you do something about them.

As for the process of quitting and transitioning to the next stage of your life, if that is indeed what you decide to do, I might make sure that you have something else lined up before you jump ship. Remember it could be a while before you actually find a job you're happy with, or even a job at all. Remaining a student while you look will, at minimum keep you on a stipend and avoid any gaps in your resume. And if that time drags out (and you address your concerns about your research), you might actually end up close enough to finishing that it's worth your while to keep going.

 

[radium]

Machine learning is a very hot topic among physicists these days. I know some theorists who left physics to do things related to machine learning. I think knowing C++ opens a lot of those possibilities. They seem to really enjoy their jobs.

It might also be important to try to avoid any sunken cost effect here. If you were to get on track with an awesome project in the coming months, would that change your mind? Things probably won't change unless you do something about them.

I think this is very important. Grad school is a roller coaster, there are often very high highs and very low lows and sometimes not much in between, which is especially true when you are doing research. But if the highs are high enough, that's what allows you to get through the lows. If they're not and you don't see the PhD as benefitting you in the next step, that's definitely something to think about.

 

[Rika]

It's bad - 3 years of your career life are wasted so don't waste any more time. Even if you start doing research right now it's impossible to do PhD in a year so don't drag it any longer. Send your resume for programming jobs and if you can't get any switch to masters in CS and build your portfolio. I mean PhD in pen and paper won't get you far and you will need to start from scratch in IT job market either now or when you finish your PhD so doing it any longer is just a waste of time.

 

[Dr.D]

You might want to ask yourself why this has happened. Is it entirely the fault of your supervisor, or have you failed to carry the ball? A PhD is largely an independent effort, so have you done all you should in that regard?

I really suggest answering these questions before making any change. If it is something repairable, you may want to stay where you are and make the repair. If you move to a new situation with the questions unanswered, is there not a high likelihood of a repeat?

 

[Charles Link]

You finally got to take the General Relativity class this last spring. How was it? Did you like it, and was it as good as you anticipated? If you did well in it, I would not give up on a dream even if it is taking longer to materialize than you thought it would.

 

[atyy]

If your programme and advisor are still happy to have you, and you would like to continue, stick it out and see if you can get something in computational general relativity, or something that you are interested in. The middle part of a PhD can be discouraging, but usually one just needs patience, especially if the environment is supportive.

If you don't want to continue, then see if you can get a masters, and look for a job.

 

[CrysPhys]

To OP:

(1) Are you at a US university? Something doesn't sound right at all. PhD programs for experimentalists have been stretching out to 6-7 yrs, but for theoreticians are typically shorter. Yet you are entering your 4th year and haven't even started your research. Something is seriously amiss.

(2) It's not clear from your first post whether you are funded as a TA or an RA. If you are funded as an RA through grant money from your advisor, there should be some focus on your research already. If you are still funded as a TA, does that mean your advisor has no grant money, and is that part of the problem?

 

[Crass_Oscillator]

It is incredibly bizarre that you haven't done research in 4 years since starting.

Everybody I know in computation/theory was able to get a publication within a few years or less and started research immediately.

I don't think you're telling us the full story.

 

[Charles Link]

It is incredibly bizarre that you haven't done research in 4 years since starting.

Everybody I know in computation/theory was able to get a publication within a few years or less and started research immediately.

I don't think you're telling us the full story.

It is of interest what might have gotten the OP in the predicament that he is, but I don't think it should be considered as a mark against him. Graduate school is a very busy time, and for whatever reason, things haven't fallen into place yet for him.

 

[Crass_Oscillator]

I'm not holding it against the OP, but his situation doesn't add up or make any sense.

 

[Charles Link]

I'm not holding it against the OP, but his situation doesn't add up or make any sense.

We need to hear from the OP, but one question is if the OP is still in good standing at the university in that did he do a reasonably good job as a T.A. and in his other studies, so that he has the respect of the department, etc. If that is not the case, it may be in his best interest to look elsewhere for future prospects.

 

[Crass_Oscillator]

One approach to these problems is to consider switching programs. I knew a guy at a low tier physics program who retook the PGRE and got a 4.0 in the grad classes. Nobody at the school did GR (in particular lensing), so he transferred to UPenn which is a pretty fancy school.

OP, you seem very disillusioned due to roadblocks and headwinds towards becoming a scientist. Perhaps the solution is not to give up, but to look for other ways to pursue that science.

EDIT: I started a PhD in computational biophysics. I utterly abhorred it and the dreadful program I was at (which had a very high reputation). I dropped out and switched subjects entirely to solid state device physics. I'm significantly happier. It's been bumpy but it's doable. The overall cost in time for me was just a couple years, whereas for you it's 4, but if you're really intent on doing science you should not give up, unless you're sure that you'd rather be a data scientist/programmer/engineer.

 

[TomServo]

You finally got to take the General Relativity class this last spring. How was it? Did you like it, and was it as good as you anticipated? If you did well in it, I would not give up on a dream even if it is taking longer to materialize than you thought it would.

I loved it actually, and I would like more of it. I worry however that TA duties will just slow things down even more even if the research side really picks up.

 

[TomServo]

Let me explain what I mean when I say I haven't done any research. I've read papers, attended meetings, ran meetings, but I myself have not sat down and done any actual research. Most of my time was taken up with required classes (and we had a debacle my first year that made things much more difficult), taking a difficult elective that in retrospect was a waste of my time, TA duties, and like I said attending and even running some meetings. But does merely reading papers and going to meetings count as research?

 

[TomServo]

To OP:

(1) Are you at a US university? Something doesn't sound right at all. PhD programs for experimentalists have been stretching out to 6-7 yrs, but for theoreticians are typically shorter. Yet you are entering your 4th year and haven't even started your research. Something is seriously amiss.

(2) It's not clear from your first post whether you are funded as a TA or an RA. If you are funded as an RA through grant money from your advisor, there should be some focus on your research already. If you are still funded as a TA, does that mean your advisor has no grant money, and is that part of the problem?

He does get some grant money but none that covers an RA. I've been a TA the whole time. Classmates of mine with other experimentalist advisors have TA positions, go to CERN, present their research abroad, etc. One is even graduating soon. Meanwhile I was waiting to take GR before I could even think of really doing research in gravity.

 

[TomServo]

It is incredibly bizarre that you haven't done research in 4 years since starting.

Everybody I know in computation/theory was able to get a publication within a few years or less and started research immediately.

I don't think you're telling us the full story.

As I said in a reply above I was attending meetings and even running some, but I had to wait until I took GR (in my THIRD YEAR) before I could really start research, and even then I'm still taking a cosmology class now. Between doing the reading and homework for the class and TA duties and no research topic given yet (we were kinda looking at one semi-experimental area but have now switched to a more theoretical one that my advisor thinks I'll be able to do a lot of work in quickly) I don't know if I'd call what I did "research." Is it?

 

[TomServo]

We need to hear from the OP, but one question is if the OP is still in good standing at the university in that did he do a reasonably good job as a T.A. and in his other studies, so that he has the respect of the department, etc. If that is not the case, it may be in his best interest to look elsewhere for future prospects.

I am still in good standing, do well at my TA duties and do well in my classes, especially GR and cosmology (which I'm taking now).

 

[Charles Link]

I am still in good standing, do well at my TA duties and do well in my classes, especially GR and cosmology (which I'm taking now).

The courses that you have had as a graduate student other than GR could one day be quite useful to you even it presently might seem like a lot of detail in subjects unrelated to GR. Being a TA does get you funding and pays the bills. It's good to hear that you've had some success in your general relativity class. I'm presently retired, but looking back on my college days, I would say one thing that was rather unnecessary was for everyone to be in such a hurry all the time. Be patient with things, and it sounds like things could fall into place for you, perhaps even quicker than you would expect.

 

[Dr. Courtney]

As I said in a reply above I was attending meetings and even running some, but I had to wait until I took GR (in my THIRD YEAR) before I could really start research, and even then I'm still taking a cosmology class now.

I published my first five papers in theoretical atomic physics before I took my first class in atomic physics. Learned it on the fly and just jumped into the research. Good thing no one told me I had to finish a class first. It was expected that we would jump in, figure it out, immerse ourselves in papers, books, discussions, start slinging code, contributing to the computations. Next thing I knew I had results to publish ...

 

[radium]

I agree that taking classes related to your subfield is incredibly useful for research. However, there is eventually a point when you need to be able to learn topics by yourself as you need them as a theorist. While I have taken classes that helped in my research, they only helped so far as being able to understand the basics. The rest I filled in myself by reading papers, reviews, textbooks (the more specialized ones can be very good for learning about specialized topics like AdS/CFT for example), and talking to other grad students and post docs.

I've also learned that sometimes you just need to start something even if you don't fully understand the big picture because you will learn along the way.

 

[TomServo]

I published my first five papers in theoretical atomic physics before I took my first class in atomic physics. Learned it on the fly and just jumped into the research. Good thing no one told me I had to finish a class first. It was expected that we would jump in, figure it out, immerse ourselves in papers, books, discussions, start slinging code, contributing to the computations. Next thing I knew I had results to publish ...

Isn't that different though because you would have taken QM?

 

[Dr. Courtney]

Isn't that different though because you would have taken QM?

If it were that easy, one would think any grad student could publish in atomic theory as soon as they completed QM. Why not give it a try and see how it goes? It took me many hundreds of hours of independent study in atomic physics above and beyond my QM coursework to be able to do work in the field worthy of publication. Have a look at these two papers and consider how well your QM courses would have prepared you to author them:

https://journals.aps.org/pra/abstract/10.1103/PhysRevA.53.178
https://journals.aps.org/pra/abstract/10.1103/PhysRevA.51.4558

I expect there may be a difference in degree, but not in kind. If you want to publish meaningful work in GR, learn enough GR to do it. No need to wait to complete coursework. Other fields I've published in without/before having a course in the field: astrophysics, traumatic brain injury, blast physics, fisheries science, physics education, and fluid dynamics. Figure out how to learn what you need to do productive work in a field without needing coursework. I had a good library and lots of great books and papers. Today, grad students have the near infinite resources of the internet.

 

[Locrian]

An interview with Matthias Bussonnier, PyDev of the week at pythonlibrary.org.

He has a PhD in biophysics, used Python in getting his PhD, and is a core developer of Jupyter, which is pretty awesome.

 

[Choppy]

If you want to publish meaningful work in GR, learn enough GR to do it. No need to wait to complete coursework.

I think I agree with the sentiment behind what Dr. Courtney is saying here, but I think a caveat may be warranted. Some people might read a statement like this and question the need for formal education entirely.

On one hand many students can experience a kind of learned academic impotence - a feeling that they will only be able to do meaningful research in a field if the "golden hoops" are jumped through. And I think that's what Dr. Courtney is rallying against. It's critically important to be able to jump in and do the background reading on your own, generate your own ideas, pursue them, read how other people have approached your specific problem, etc. Because formal hand-holding is only going to get you so far.

But on the other hand, it's also important to recognize the value of good mentorship, building a solid foundation in physics in general, and learning from high quality sources. The internet is a double-edged sword, which while allowing unprecedented access to information, is also filled with crackpots whose personal theories have "proved Einstein wrong," who are the result of the pendulum having swung too far to the other side of the spectrum.

It sounds like prior to taking this course, the OP was doing a lot of reading. In my opinion that counts as research. One of the other dangers of jumping in too soon without guidance or having done enough background reading is that you risk doing a lot of work on a problem that's already been solved. Reading isn't going to earn you the PhD, of course, but it is critically important to it.

 

[Dr. Courtney]

I think I agree with the sentiment behind what Dr. Courtney is saying here, but I think a caveat may be warranted. Some people might read a statement like this and question the need for formal education entirely.

Thanks for the opportunity to clarify. Not at all what I meant. I'm speaking in the context of someone who has completed a BS in Physics and is in a PhD program AND in a research group in GR.

I was only able to do research in theoretical atomic physics in an analogous situation. My adviser (Dan Kleppner) was (and is) a leading light in atomic physics. He was available every day for students to go into his office, get advice, and bounce research ideas off of. I'd go in about once a week and he'd steer me away from rabbit trails and unproductive paths. Dave Pritchard and Wolfgang Ketterle were just down the hall and all the grad students were (more or less) a big fraternity helping each other figure things out. We met with most of the big theorists in the field on a regular basis AND these guys (and/or their grad students) would answer emails and always help you out to share ideas/code/brainstorms and keep projects moving along.

The idea of a lone wolf doing real work in serious theory is not what I contemplated. But with real support, there is no need to wait to complete _coursework_ in a specific field before starting research.