My PhD Journey: Disappointment, Hard Work and No Results

[cosmogirl]

Hello everybody!

I'm now in the second half of my PhD program, with 2 more years to go, and nothing to show for it. Sure, I've written two papers (which got zero citations), but the only good thing I can say about these papers, is that they are not the worst papers ever written. Unfortunately, my adviser wants me to continue pursuing this line of research, because he just got some research grant for it.

When this first started bothering me, some time ago, I was depressed for a while, and then decided to take some action. So I started 2 other projects, which seemed (at the time) more interesting and more promising. This decision, and a coincidental increase in my teaching load made last semester the most difficult one in my entire life. Then I continued to work very hard all summer, hoping that I would soon begin to see some results.

However, things didn't work out quite as I planned. What I have now on my hands is three projects which are still far from completion, none of which I can dismiss because I'm tied to other people, and zero motivation to continue. As I see it now, these projects will at best produce more uninteresting papers that no one will ever read, and even that will require tremendous effort on my side.

By the time I finish my PhD, not only I won't have any chance of getting a postdoctoral position, but (more importantly) I will have wasted all my time in grad school in vain. I feel that all the work I've done is completely useless, nobody will ever use my results or my code. Not that I hoped to make any groundbreaking discoveries, but the work I've done so far is of no use whatsoever. This situation is very depressing, to the point that I can't make myself do any work at all.

If anybody was in this situation (or their friends, maybe) I would really like to hear how they managed to pull themselves out of it. And general comments and ideas are also highly appreciated!

 

[mal4mac]

Find a job, I jumped ship and never looked back! You mention "code", so I'd expect programming jobs to be an option. I took well paid programming jobs that looked interesting, and there was no problem moving on if they weren't interesting.

 

[cosmogirl]

Thanks for your reply, mal4mac!
The problem is, I have to finish this PhD anyway. Also, I enjoy doing research, I just don't enjoy the part where you're stuck on something for several months with no real progress in sight... And when I look around I see other PhDs are doing some useful and creative work, that pays off right away.
Maybe I'm not good at doing research and a "real-world" job would suit me better. I'll still have to finish my program to find that out, though.
Did you switch to a non-academic job in the middle of your PhD?

 

[chiro]

I agree with mal4mac: if you have done a substantial amount of coding where you have the pains of going through massive amounts of code, fixing arcane bugs, using lots of 3rd party libraries (both with the code included and without) and contributed to a large repository getting the experience of breaking it and all the fun that this experience brings, then I think you have a chance to do dev work.

You will have to know the language very well because a technical interviewer will ask really obtuse questions just to get a sense of the depth of your knowledge (I remember in one interview I got asked what the mutable key word was in C++).

Also completing something is really important (possibly more important than you realize).

One of the biggest risks employers take is getting someone that won't "stick it out": it's not about getting the best person for the job, but rather getting someone that can stick it out and get something done with a result that is adequate enough.

Showing that you have completed something in terms of x years shows commitment and that's a massive thing because people do get tired, burn out, emotional, and things go wrong and the last thing a project needs are people that become so demoralized that they give up half-way through (or earlier) so when it comes to rule of thumb evaluation, if they see you can do a project for five years and see it through, then that will be an indicator to them that you aren't a flake.

I don't know what kinds of things you want to do, but that one attribute is really good to have especially in project industries of which one is software development (also for long term consulting as well).

 

[mal4mac]

My situation was in the UK in the 1980s, so things might be a bit different in your circumstances. I took an MSc and therefore could only get two years grant for a PhD with a vague suggestion that some extra funding might be found. A year and a half in one of my supervisors had stolen my most promising line of research and was coding like crazy to get papers out with his name first. I complained to my other supervisor and he said I should be overtaking him doing the same coding - which I thought was mad advice The other guy was a world authority in the kind of coding we were doing, with twenty years experience... so I thought - no money looming - bad supervision - "time to get out." I easily landed a well paid research assistant post at another university.

The RA post was OK for two years but, I was starting to be unhappy with the way I was being treated (again!) For instance, the prof had published papers without my name on it, when I at least deserved "second name"; plus some other issues. So I side stepped into a teaching assistant post, quite fun & useful as I learned a lot of CS stuff and got paid for it, which looked good on my CV.

From that point on it was a doddle landing just about anything I applied for, mostly research programming jobs in the University sector. So there's very much life after ditching a PhD!

Why do you "have to" finish this PhD?

You have published two papers so you are "by definition" good at research. So what if you haven't been cited? For a start, you only published them recently!

P.S. Stop volunteering for projects, unless they pay you consultancy rates (!) I used to do that, then I realized I didn't have to... just say you already have too much work on your plate. Even if you don't...

P.P.S. I never had it so hard Chiro - in applying for I-don't-know-how many jobs I didn't have a grilling from a hard tech guy. (I *was* the hard tech guy... I'd several tyars experience in Algol and Fortran and that got me into jobs in Visual Basic, Object Pascal, "teaching", Objective C, Smalltalk... no hard tech guy... just relaxed profs saying things like, "You can program in Algol? Picking up Visual basic [or whatever] should be no problem then...", maybe it was more relaxed in UK Universities in the 80s?

Note, I didn't complete my PhD, so non-completion is far from a career ender... many computer projects fail, many PhD people drop out. These people usually get good jobs & fun times afterwards. Cosmogirl, in publishing two papers you have completed (twice), I had my name on two papers, as well, at the time I dumped my PhD - not first author though. And that was good enough...

 

[homeomorphic]

I'm in a similar situation, except quite a bit worse.

I have no papers and I am in my seventh year, although in pure math, it's common not to publish much until later in the game. Two useless papers is better than no papers. Plus, I don't know how you can know that something is useless.

Today, I had a major setback in my thesis and realized that one of my theorems that I was trying to prove is not true at least as stated. I don't have that much time left, so it's hard to recover from any disasters like this. So, my PhD is up in the air now, just as I recovered my will to try to finish. The problem is that I thought the problem I was working on was easy, but it turned out to be much harder than I thought. I didn't realize it until too late, so I didn't take it that seriously. Now, it may be too late.

I haven't pulled myself out of the situation. I would say it's better to finish if you can, but you shouldn't put too much pressure on yourself. It's not the end of the world if you don't finish.

 

[mal4mac]

Seventh year?! What do you do for money? What are you planning to do next?

 

[homeomorphic]

I have had funding up until now, doing TA work. Not sure about next semester. I have a little bit of money saved up, but my budget is already tight and I might have to do some tutoring to get by if I am not funded next semester.

As it turns out, there appears to be an easy fix to the problem that I didn't see due to a mistake that I discovered yesterday, so it looks like I will still be able to graduate at the moment.

 

[bcrowell]

Most research is uninteresting and unimportant, even to people within that particular subfield. It's actually a good thing that you're able to look at your two papers you've done so far and realize that they're uninteresting. It shows objectivity and maturity.

In the course of writing these two papers, you've presumably gained a lot of expertise in a particular field of research. This is a good thing. You can't make important, exciting, original contributions unless you first develop a certain level of competence and familiarity with the field.

You might want to tell us what type of research you're doing. Different fields are very, very different. Are you working basically alone, in a small research group, in a huge collaboration ...? Experiment? Theory?

Option A would be to stay with your current adviser, explain your concerns to him/her, and not let yourself feel compelled by various interpersonal commitments to work on topics that you find uninteresting or that are not working out. Honestly, I don't think anyone is going to hold a grad student's feet to the fire if she says she simply doesn't think a particular topic is a good one.

Option B would be to switch advisers and switch to a different subfield or field of research. This might entail adding another 2-5 years or so to your time in your PhD program. I knew several people in grad school who did this. Your ability to do this might depend on your funding situation. You could treat this as a backup option in case your adviser nixes option A.

I'm confused by your statement that you have "2 more years to go." PhD programs aren't on a clock, especially for experimentalists -- the experiment is done when it's done.

 

[cosmogirl]

Thanks everybody for your helpful comments!

mal4mac, I'm not in the UK, but I think everywhere things are very different now from what you describe. People who finish their PhD and want to find a programming job spend a few months looking for one and going through a grueling interviewing process. Most of them say their job is not as interesting as their research was, but I think they get to see the result of their work every day, or at least every month, as opposed to once a year in my case. :)
On the other hand, I've always wanted to do research in physics, so while it's nice to know I have other options, it would be very disappointing to give it up in the middle of my PhD. So I guess I'm trying to find ways to rekindle my passion in physics, so to speak :) and finish the projects I'm working on now.

homeomorphic, it seems we're in similar situations. I think it's better to finish a PhD in any case, if only to not have a feeling of missing an opportunity a few years down the road. Also, it seems (as mentioned many times on this forum) that potential employers value the fact that you were able to take a long term project and finish it. That said, I still have two years of funding, I don't know what I would do if I ran out of money. Good luck to both of us! :)

bcrowell, what I mean is that I have funding for two more years. My research is in theory, and I'm working mostly alone. Maybe I would feel better if I worked in a larger research group, I don't know.
I learned a lot from writing those two papers, no question about it, but I guess I expected more from myself. Also, the fact the nobody cited them, hence nobody used my work, is a little depressing.
Of course it's a good idea to talk to my adviser, I'm not sure he'll help me out, though. I think the best option would be if I could somehow finish the current project (one of them at least) and then to work on something completely different. If only my adviser would agree, I think I could live with that.

But do you really think that most research is uninteresting and unimportant? Incremental - yes, but even the tiniest building block is important if somebody used it as part of a large structure.

 

[I like Serena]

Perhaps you should try to share your research with someone else.
A colleague, a friend, an ex-PhD'er, your advisor, another advisor, ... anyone.
It is depressing to do everything on your own.
I believe feedback of any kind is a prime motivator.

 

[bcrowell]

But do you really think that most research is uninteresting and unimportant?

Yes, I really think so. See, e.g., http://chronicle.com/article/We-Must-Stop-the-Avalanche-of/65890/ Only about 40% of science papers are ever cited.

If your work was wrong or incompetent, that would be a reason to beat yourself up. If your work wasn't good enough to be published, ditto. The fact that it hasn't been cited simply shows that you, as an apprentice researcher, are doing about as well as most people in academia.

The two years you have left before you get your PhD are supposed to be years in which you grow as a researcher. (If you weren't supposed to grow between now and then, why wouldn't they already have given you your sheepskin?) You've built your basic level of competence, but, to your credit, you're not satisfied with publishing competent but unimportant work. That's what those remaining two years are for: they're for you to keep progressing to the level of originality and noteworthiness that you aspire to in your work.

 

[mal4mac]

I have had funding up until now, doing TA work. Not sure about next semester. I have a little bit of money saved up, but my budget is already tight and I might have to do some tutoring to get by if I am not funded next semester.

As it turns out, there appears to be an easy fix to the problem that I didn't see due to a mistake that I discovered yesterday, so it looks like I will still be able to graduate at the moment.

Good stuff. But after seven years are you not tempted to get a job? This isn't a veiled criticism, I think it's great *if* you have joyfully followed an intrinsic interest for seven years, even though living on a pittance.

After two years I got bored with the Math I was using in my PhD research... then they had to pay me to do other Math, which I found quite interesting... nice when you get the cake and eat it :)

 

[mal4mac]

Option B would be to switch advisers and switch to a different subfield or field of research. This might entail adding another 2-5 years or so to your time in your PhD program. I knew several people in grad school who did this. Your ability to do this might depend on your funding situation. You could treat this as a backup option in case your adviser nixes option A.

Five more years living on a pittance? In the UK, I would say go for a Research Assistant post that gives you the possibility of taking a PhD (if the PhD is a must for you...) At least you would then have a wage that was similar to that of a school teacher. Such posts also have set hours so you can work 9-5, and start to live a reasonably balanced life...

Or is it possible in the USA nowadays to get a PhD studentship equivalent to a good wage?

Physics is supported by the military-industrial complex because it makes Big Bucks for the fat cats at the "top". You should at least get a reasonable slice of the fat cat cake!

 

[f95toli]

Also, the fact the nobody cited them, hence nobody used my work, is a little depressing.

The fact that nobody cited them does not mean that no one read them! You don't cite all the papers you've read when you are writing a paper, you usually stick to citing the most imporant ones that support whatever it is you are doing. Moreoever, most journals have length constraints, meaning you really try to keep the bibliography short.

I cite perhaps 10-20% of the papers I read, that does not mean that I don't learn anything from reading the other 80%.

 

[cosmogirl]

i like Serena: Certainly, working alone is difficult. Of course, I share my work with my adviser, but perhaps not enough. Usually I like to talk to him when I have some kind of a clear picture in my mind, or, better yet, some results. When everything is messy I try to avoid him :) probably not the best idea. The problem is, when I'm stuck, like now, I feel guilty about it, and so don't especially want to talk to my adviser at times like that.

homeomorphic, I know it sound weird especially after what I just wrote about myself, but maybe you, too, should talk more to your adviser? He should know better whether the problem you're working on is easy or not, and how much time it should take you to solve. Of course I don't know if you can tell that in pure math.

f95toli, that's actually a comforting thought! I, too, read tons of papers and only cite a few. Of course it would be better if somebody used my results in a more direct way... After all, this is the purpose of doing research.

mal4mac, the PhD studentship is usually quite enough for a comfortable existence (I guess it depends on where you live, etc.), especially if you're single. Otherwise we wouldn't have that many PhDs...

bcrowell, that is a scary article! Maybe it exaggerates a little bit to make the point, but the numbers are appalling anyway. I'm not a big fan of the publish or perish system, but I'm not in a position to change the rules of the game. I would prefer to work on longer-term projects that are more interesting. Instead, I work on things that seem to be short-term, and end up wasting a lot of time anyway. I do hope to do some worthy and interesting research so I want to continue my PhD, and you're right - I still have these two years, I only hope I'll be able to use them wisely. Thanks for the support!

 

[homeomorphic]

homeomorphic, I know it sound weird especially after what I just wrote about myself, but maybe you, too, should talk more to your adviser? He should know better whether the problem you're working on is easy or not, and how much time it should take you to solve. Of course I don't know if you can tell that in pure math.

Essentially, I already know how to solve the problem, I would say as of Sunday when I worked out how to fix my mistakes. Assuming everything goes smoothly and no more mistakes are found, the difficulty is to proceed with the rest of the plan and write down all the details. I have spent months trying to fix a certain technical lemma that I have alternating thought was either proven or fixed the whole time. I am writing up the final solution, as we speak, with Latex open in another window, keeping my fingers crossed that no mistakes are found as I do. Once that is done, the rest of the thesis appears not to pose any serious difficulties, other than just getting myself to write the stupid thing down and then do all the editing, which will take a huge amount of work. My adviser says he doesn't see any mathematical difficulties.

 

[zapz]

Fingers crossed for you homeomorphic!

 

[atyy]

Yes, I really think so. See, e.g., http://chronicle.com/article/We-Must-Stop-the-Avalanche-of/65890/ Only about 40% of science papers are ever cited.

I agree with your point that most publications are at best not important. However, short term academic citations are not the only way a paper should be judged. Take the case of Chowning, who was denied tenure, but whose work eventually earned Stanford big bucks.

http://www.wired.com/wired/archive/2.03/waveguides_pr.html
http://news.stanford.edu/pr/97/970709sondiusxg.html
"Stanford's FM synthesis patent, which expired two years ago, was the second biggest money maker in campus history. It brought in more than $20 million."