I am currently in a problem-solving mode of my PhD (theory, middle stage). I am working on something which is incremental on the last half a year. Until now, I did not get anything non-trivial. I noticed that I have become slow as compared to the initial two years of my PhD.

I am wondering how to speed up my research. I have tried to scan as many papers as possible. I am also trying to discuss my research problem with other students, but it is not working.

Question: Is it natural that you become slower in the middle of your PhD? How can I speed up my research?

  • 19
    Every phd student I have known (including myself) has panicked in their second year (middle of a three-year phd typical in Germany) that they don't make sufficient progress. Most of them have then been very productive in their last year. I'd say your experience is not unusual. – Roland May 16 at 12:11
  • 1
    It takes time to mature and for the material to take root in your brain. Sometimes the fruits / insights of your studies come directly, but sometimes knocking anytime 1-10 years after you first took the course / read the book / had that talk with a professor is not too uncommon. It is a natural process. The only thing you can do is to keep at it and not give up. – mathreadler May 16 at 18:22
  • Sidetracked discussion about paper-output statistics moved to chat. Please continue discussing there. – Wrzlprmft May 17 at 7:56
  • Tweet from a PhD procrastination expert (literally his career). You are in the bottom part of that rollercoaster, as most PhD students have been, are and will be. – Ander Biguri May 17 at 9:22
up vote 29 down vote accepted

In many disciplines, research amounts to an endless series of frustrations, setbacks, and failures punctuated by the occasional success. So while it’s not necessarily true that there’s a slowdown in the middle, it’s usually the case that there are periods of lower output than others.

The only projects that go smoothly from start to finish have either been extremely lucky or tend to be development-based or “following on” from someone else’s struggles. So don’t be discouraged by having a period of relatively fallow productivity—it happens to nearly everybody. It’s a sign that your problem is non-trivial and requires some work and effort to complete successfully.

As for “speeding it up,” I don’t think that’s really possible to do by scanning papers instead of reading them. It’s hard to make specific suggestions without knowing more about the discipline, but productivity in research can be elusive at times, and there really isn’t anything you can do to accelerate things, unless you can somehow try more different solutions during your workday. So perhaps it’s a matter of having fewer distractions and disruptions to your day.

  • 3
    Are there actually disciplines where research is not an endless series of frustrations, setbacks, and failures punctuated by the occasional success??! – JeffE May 16 at 15:44
  • 10
    @JeffE Yes, generally very new disciplines, which are often ignorant that they are repeating the same things that were done in another field 50+ years ago. – ttbek May 16 at 16:22

As some one who's done their PhD in 3 years and supervised PhDs for 10 years:

1 - Cut the crap

It's easy to get overwhelmed with all the papers you read, so many topics, so many possible directions to go in. You need to focus. I suggest you start your first paper right now. From all the research you've done, what's a small topic where you can contribute? Make a list. Pick one. Start there, decide what more research you need to do to turn it into a paper, do the research, write it, publish it.

I did exactly that after 1.5 years in, and really just by fluke. My paper wasn't even exactly aligned with my main research topic, yet it did wonders in getting me going.

2 - Chunk it

Once you've written you first paper, keep going. If you get to five published papers, you're done. And the great thing is that you have your PhD already mostly written, just fill in the gaps, thesis done. It may seem counter-intuitive as you want to answer the one big question, but trust me, once you chunk your big research into little topics it becomes a lot easier. And you figure it out along the path. Build on each paper, you'll get there. You may even not get to answering your original research question fully, but with 5 published papers, who's going to argue with you?

3 - Ask the experts

Talk to other lecturers, researchers even outside your faculty where it fits. I had some problems I couldn't solve. Turned out most of the experts in the field at my uni hadn't got a clue either. Until I found one who did. Go to conferences, make connections. That'll also help you in your post-PhD life. Don't give up, someone out there will be able to contribute something quite valuable.

4 - "Trust your instincts, Luke"

The brain works in mysterious ways. I chewed on a problem for a year. Then, one morning in the shower I solved it. This is just how it happens. There's a physicist who said, "men who don't shave can't solve problems." Times of self-reflection, that's when the ideas come. If you could figure the answer out in a day, they wouldn't give you three years. And I'd fit in exercise, even just go for an half an hour walk. Lot's of famous people got their ideas that way.

5 - Get off the hamster wheel

I certainly didn't work hard, yet did it in 3 years. There are students working 14h days including weekends. And they take 5 years. Why? Because all their doing is running on a hamster wheel. I had a student who went kite surfing every afternoon and did a great thesis in 3 years. Why? Because he focused on the finish line. It's easy to get caught on the hamster wheel. You're doing work, right? I spent 14h in the lab today, well done, pat on the shoulder. The reality is, if it's not helping you to get to the finish, you're wasting your time. Stop self-gratifying work, do the hard stuff, the work that'll actually get you there. And if you don't know what that is, figure it out. Yes, it's harder than just sitting down and doing "something". Take a snapshot of what you know and make a plan from that. Keep checking your progress against the plan and modify the plan as necessary. And that links back right to point 2 - if you keep pushing out papers, you're making measurable progress. Plus, you build confidence that what you're doing is actually valuable. You got proof in print.

I would see a PhD studenship as a superimposition of two effects.

  1. The more you know and the "deeper" you are in your topic, the more efficient are you.
  2. The more projects, topics, contacts (!) you have, the slower you are.

The second issue is present in every job I experienced, my experience being limited to programming or academic jobs.


Why the preface? Well, in your second year you get faster due to (1), but the back-flow of (2) makes a significant appearance. Producing a revision for a previously submitted paper, fleshing out a small improvement in old code, understanding old code or someone's else research – all these required issues slow you back, sort of. Of course, then you can boast another well-placed paper, know more about some further methods, etc. But at the moment it's perceived as a slow-down.

In your first PhD year you can just start reading / programming / whatever and it feels super productive, even if the actual research output is lower.


tl;dr: Don't fret, it's a phase.

I can't say that your experience seems any different from what I went through in my PhD. Completing a PhD is intentionally difficult. There is no magic formula to completing a PhD, no quick fix or way of speeding up the process- in fact, if anything, panicking and trying to find fixes could potentially have an adverse effect on your mindset impacting how quickly you complete your research.

Scanning papers will definitely not help you. However, pick a particular thread and pull it. What I mean by that is thoroughly read a couple of papers and then work back through the references they provide- really deeply get involved in one particular aspect and see how that relates to everything you do. There is no point using a scattergun approach- the only effect that will have is to make you think you're making progress when really you're probably not!

I left a PhD due to some extreme life hardship (unrelated to the PhD), things I wouldn't wish on anybody, but going through really rough times taught me a lot about how to be successful. Here is my advice.

1. Personal development is not wasted time

It may be tempting to spend all your time on your PhD. But even seemingly frivolous pursuits like learning to speak a new language or learning neuroanatomy will help you find new solutions to problems. Read books or take cheap online courses related to productivity, creativity and success.

2. Learn to think for yourself

It's easy to run a test and discover a 30% speedup for a 10% cost. It's a lot harder to answer the question "What proportion of businesses will be willing to spend money to implement this?" or "How come nobody is using my tool?" To answer these questions you need to be able to put yourself in someone else's position; you need to think about the dynamics of the communities and publication channels through which people learn about new things, and you need to try to reverse-engineer the success of other ideas. For instance, I discovered a powerful library for my research because a student had written a series of "novice-friendly" blog posts about how to use the library his group had developed, and this came up in Google while I was searching for something else.

This kind of thinking will help you pursue ideas with more value and impact. I have found that as I got better at this, I became much better at coming up with clever ideas, too.

3. Go back to first principles

There are a lot of implicit assumptions and habits of thought in the way we currently do things. But often they are imperfect. They are useful tools, good approximations, but don't take them as gospel. Explore the unwritten assumptions. Get your head around why these tools are effective, but also try to understand their weaknesses and limitations. Think about problems from first principles, without the support structure. Think about the problem in a "boots on the ground" perspective instead of an "ivory tower" perspective. (Depending on your field that may be rushed doctors needing to make "least risk" judgments, or the crazy compiler and CPU optimisations and cache hierarchies that make actual computer performance quite different from the theoretical.) Two perfect examples are asymmetric cryptography and cryptocurrencies, which both solved a theoretically impossible problem by using solutions that are theoretically breakable but are good enough in practice.

4. Valuable habits often sound crazy

I found that simply writing a weekly report on my work was often a surprising source of new ideas. But doing unusual things can often spark even more creativity. Here are some ideas:

  • Hand-write or dictate a report instead of typing it.
  • Write a "letter your grandmother could understand" once a week about your work.
  • Write an article once a month blasting your work (i.e. tearing it to pieces) or religiously defending the current state of the art.
  • Write a list of the problems you are currently stuck on. Create a visual representation of the problem. Now simplify each problem to a "logo". Choose a piece of music that represents how you feel about each problem. Listen to them in random order.
  • Do a similar thing for existing solutions to related problems.
  • Use the Po technique to explore solutions to each problem. Try other de Bono techniques.
  • Record yourself giving a short (or long) overview of your research topic. Play it back and hopefully things you missed will jump out at you.
  • Create mind maps of problems and solutions. Try finger-painting them.
  • Read out a description of your research area but replace one of the key words with a random word like "pigeon".
  • Don't forget there are lots of apps that may be able to help.

Different things work for different people, of course. But try to stick with something for a while, as it is very easy to lose a good habit.

Focus. A common mistake that PhD students make, and in fact many humans make throughout any career, is that they don't focus when the time is right. In the beginning, it is very good to venture into everything and anything to see what's out there and what is worth doing. After the first year it is about time to draw a very specific goal with a sharp plan that will get you there even in the most pessimistic scenario. That's where your PhD adviser comes in most handy as they know this game very well. Do not disperse your attention on anything else until you get that PhD diploma. At some point closer to your 3rd year you may (and should) find out that there are better things to research or do in general. Good, make a quick note of it and put it away until you are done with your PhD. Until then, follow the plan and get it done.

Your Answer

 
discard

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.