PhD in a probably hopeless research project

Question

How to proceed with my PhD? I'm 18 months into it, but I don't see this project going anywhere.

Perhaps I didn't do my due diligence or researched the position before accepting it. I'm working in a team with half a dozen other PhD's, and the "research" focuses on extending and improving an extensive software system. Thus, I'm not working independently and depend on my colleagues and professor.

The professor thinks it will revolutionize our field, but that is nonsense. The few articles my colleagues have published are either conference papers or on arXiv. Much of it is equations and "fancy words" that hide the lack of novelty. I don't know if my colleagues know it or if they are cynics and don't care. Perhaps they are just here for their PhDs. To boot, the system is old, and much of my time is spent bug fixing lousy code written by PhDs who left many years ago. I feel cheated. I was promised cutting-edge research, but this is not it. Some of you will probably say that I'm not "understanding" the research. At first, I thought so too. But it's been 18 months, so I understand now that there is nothing to understand.

Everything else about the job is excellent. Both my colleagues and professor are friendly and caring people. But they are perhaps not that "sharp."

I don't know what to do. I don't want to cause any conflict and perhaps become a pariah in the office, but I don't want to waste my time. I have ideas for the research I want to do, but I don't want to offend everyone. I don't want to quit my PhD because I want to work in science.

Answer 1

If a substantial part of your research community thinks the approach "will revolutionize the field", then a sufficiently deep analysis of the reasons why it's "going nowhere" might, in itself, constitute a substantial original contribution to knowledge that gets you the PhD pass. That was essentially the nature of my thesis 20 years ago: it didn't make me popular, but it did make me a doctor.

Answer 2

It's strange that nobody has explicitly mentioned it, but I think you should start with clearly understanding your own goals. Why do you study for PhD and what are you further career goals?

Is it just get a PhD badge and switch to industry? Is it get a PhD and move to another "average" group? Get a PhD and then move to a top-level group or get a position in a top-level university? Etc.

In some of these goals (at least in the first, maybe in the second too), it does not really matter what is the quality of your research, what matters is a PhD badge. In this case, the next question to ask is very simple: what about PhD students in your group some 3-6 years older that you? Did they get their PhD? Based on what work? What do they do now? You may find that they did get a PhD with a work similar to what you are doing. In this case you have pretty good chances to get a PhD too, and that's all that matter.

In fact, sadly, in many parts of the world there are tons of really weak PhD awarded. In many places the majority of all PhDs are no novelty at all and are "hopeless" as research. Obviously, a good student should do some due diligence and avoid such places, but if you got as deep as 18 months into this program, it may be better to complete it and move on. Even such a PhD may be useful depending on your goals.

For example, I think that in industry they will not care what your research was, and moreover they will be pretty aware that there are many places like yours, and that it's not you to blame. Similarly, many "average" research groups in your city or uni may also be well aware of this, and will not judge your PhD research, but will judge your skills. So after you get a PhD, you may be able to switch to a better group and start a real research.

At the same time, if you plan to apply for some relatively top-level research group or position, then they will judge you by your work. In this case you need to get better research now. You may try discussing it with your advisor, although personally I don't think it will get any good results, or you need to switch advisor or PhD program in general.

So, think of your goals, study the careers of PhD students from your group who are 3-6 years older than you, and after that decide.

Of course, you may find that there were no PhD students in your group in recent years, or at least no successful students. That, combined with what you wrote in the question, is a very bad sign. In this case, your advisor most probably does not know what you should do to get a PhD (both formal and informal requirements), and with a very weak research you probably will not be able to figure it out yourself. In this case, run away from this advisor.

Answer 3

Many (most?) PhDs go through "second-year-blues". Essentially:

• at the beginning the PhD candidate doesn't know enough to know the problems
• then they learn what the problems are and realise that they are hard and the topic is even more difficult than they thought
• but they usually scrape something out of it by the end (which is almost never revolutionary and may in fact be "here are the problems")

Sure, there are always some people who have amazing rockstar PhDs (as with any walk of life) but the vast majority are grains of sand standing on the shoulders of giants. In a way, research is a fractal Dunning-Kruger effect. I was super pleased to discover that 1 person has read and cited my thesis, even if they were another PhD student of my supervisor.

After 18 months software engineering experience, regardless of getting a PhD or not, you're well equipped for a career in tech (science-related or otherwise). "I want to work in science" covers many, many things besides the stereotypical view of professors (which don't exist because they spend most of their time on grants or papers or teaching or administration or other "not science" activity). And if you think about it, many many PhDs graduate every year but the turnover in academics is much much lower. So a large majority of PhDs must not become professors and that's fine.

TL;DR think about what day-to-day bits you like the most; coding, maths, presenting, reading, etc. If you can do more of that and your PhD, great. If not, then that's fine too and perfectly normal.

Answer 4

It seems you only have two choices: discussing the issue with your supervisor/advisor or switching schools, both of which are risky. Try first talking with your advisor about your issue and see if you could possibly switch supervisors or get advice from them that way, or you could talk to your professor themselves, saying something like "my research interest is in ____ not ______, and this project does not work for me, so I want to switch to another professor as my supervisor". Alternatively, you could wait a year and apply next year to other universities. I don't believe they'll view your leaving mid degree as a bad thing, tho I think there's a thread on here that could give you more info on that. This time check the faculty's pages to see if they're more aligned with your interests instead of just going with whoever. Fill the time till next year by applying to grants and other awards, they look good on your CV, they'll help you get into the other schools easier, and they'll help you financially.

Answer 5

First things first: In your country, there must be laws regulating Ph.D. programs and outlining what the requirements for completing such a program are. It can go along the lines of:

• Demonstrating knowledge and application of research methods
• Demonstrating independence in research work
• Creating publishable work
• Presenting original results in domestic and international venues

Knowing these requirements can help you to gauge if your day-to-day work helps to advance towards the degree or not.

It helps to have a mentor or a broader examination board helping to gauge your progress towards said requirements.

Improving existing software and bug fixing does not sound like publishable work at all. At least not in the fields of software engineering and computer science. A novel and publishable work would be, for example, the development of a new algorithm. It would make sense to demonstrate the use of the algorithm with a simple implementation. However, anything beyond that is development work, not research.

Are there any other research groups or Ph.D. students that you can approach and gauge if you would like to work with them? If you raise any concerns, they should be about the supervision/topic falling short of the minimum bar, not your dislike of fixing bugs.

Answer 6

Every research team define specific roles and objectives for each member. Your part does not seem exciting. But the problem is regarding whether you are getting your PhD or not. I believe it is not hard to asses your situation:

1. Is there sufficient literature and articles (references) to support your project? Is it so "ground-breaking", as your professor believes? Could you prove you are contributing to the state-of-the-art in the project's corresponding field?

2. Based on those mentioned above, "few articles on conferences and arXiv" seems like a red flag. For example, my PhD program demands me to be the corresponding (first) author of two articles in world-class publications where the sum of their Q is five at most. That is, at least one Q2 and one Q3. That is not trivial since, at those quartiles, a thorough peer review process is expected to demand original contributions to the current state of knowledge. In my case, each research team member in their area should deliver two top-notch articles of original contributions as corresponding authors. Fixing bugs and a 5th or 6th place in the author's list won't get me anywhere. Is that your case in your PhD program?

Answer 7

"Second-year-blues" quoted by @afaulconbridge is the first thought I have when reading your situation. However, pointless projects also exist. An external input from a mentor not involved in your research group and not connected with your PhD, but knowking computer science and the academic world, would probably be highly welcome to better understand your situation.