18

I feel that I can get into big trouble very soon. I have been working on my master's thesis in computer science area for about 3 months(out of 6) and still haven't achieved any meaningful result. Although I am working really hard(at least 8 hours per day + some time during weekends), I can't see any hope how to complete this research in remaining 3 months and compile it as a thesis.

The problem is that my supervisor selected very ambitious project for me. Initially, I was a bit scared, but general idea was looking quite meaningful. He promised that "we will be the first who have done this, bla-bla-bla....". In addition to this, he also offered quite generous funding. And I swallowed the bait)

Now, after several months of research, I can claim that the method we are developing looks a bit meaningless(that is why there are no any research on it). Maybe if we dig deeper, this will start making sense, but MS thesis is just for 6 months, I cannot afford doing this research for 1-1.5 years.

So now I am considering dropping out of the project and finding a new supervisor. Would you recommend doing this? Or, otherwise, how to deal with MS research which yields no results? How this can affect my grade?

  • 22
    What did your advisor say when you told him about your progress on the project to date? (You do give your advisor status updates, right?) – Mad Jack Jan 28 '18 at 0:34
  • 7
    "Not beating the state-of-the-art" != "Not achieving any meaningful result". That can be respectable. Especially as this is only an MS, and 6 months at that, not a PhD. – smci Jan 28 '18 at 17:03
  • 4
    You got funds for a MSc thesis? :o – Andrea Lazzarotto Jan 28 '18 at 17:53
  • There are 6 months allocated for your thesis. Try to stick to that. If the initial project is too ambitious, try to deliver something less ambitious that still adds value. That is: somebody else can pick up and continue your work. Delivering in time is a quality that is worth a lot. – jos Jan 29 '18 at 10:08
32

Your situation is quite common with MS theses (in the US at least). There are several people who are given ideas from their advisors which either turn out to be dead-ends or take up much more than the allotted time. In EECS, the MS thesis defense is more about displaying your ability to conduct research than having significant results. Clearly describe the problem, your methods, and display preliminary results whether they are significant or not.

As an example, a professor in my university is interested in inverse problems in imaging. Every once in a while he takes up a student for an MS thesis for denoising images. Problem is, nothing really beats an algorithm called BM3D in terms of PSNR for denoising images (except in specific circumstances). So he has had a few MS students who developed a new method which didn't work as well as BM3D but nonetheless conducted meaningful research and successfully defended. Everyone on your committee knows that it is impossible to anticipate whether an idea will actually work before doing it.

Switching topics and advisor halfway through might not work (how do you know that a new idea is doable in time before actually doing it?). Usually, this is best reserved for situations where there are inter-personal conflicts. Assuming that your advisor is looking out for your best interests, schedule a meeting and talk to them about your concerns with the amount of time you have left and what aspect of the project you should focus on.

  • 5
    I had to look up what "EECS" was about. For future readers: apparently, some universities mix up electrical engineering and computer science, two vastly different fields which require different skills and mindsets, under the same umbrella. – Andrea Lazzarotto Jan 28 '18 at 17:56
  • 12
    @AndreaLazzarotto Mix up? EECS departments exist at e.g. MIT or Berkeley. There are two fields that CS (mostly) emerged from, mathematics and electrical engineering. Did you not have a lecture on e.g. computer architecture during your CS studies? Did you never see a transistor, logic gate, ALU, register or anything during a CS program? What are these things if not electrically engineered components? – Adrian Jan 28 '18 at 18:35
  • 5
    @AndreaLazzarotto Well you might be surprised how many researchers in the CS filed hold an "engineering" degree. Many of my CS professors are "engineers" according to their titles, since their PhD theses were rather about applied/practical aspects than purely theoretical ones. If you have sb in the field of software engineering (sic!), computer networks etc. these people are actual "engineers" in the whole sense of it. – Adrian Jan 28 '18 at 18:50
  • 8
    @AndreaLazzarotto You're mixing up "enginner" as a trade profession and as somebody in the academic field. People who do plumbing can get an "engineer" certification in some countries. It has nothing to do with engineering in an academic sense. And regarding to your link, my German professors have the "Dipl.Ing." title that your wiki article states. You can get a "Master of Engineering" in many CS courses as well. – Adrian Jan 28 '18 at 19:44
  • 1
    @Adrian, you keep switching topic to avoid the fact that engineering and science are two different (although partially overlapping) fields and engineers and scientists have different academic paths. It's fine, I am not interested in arguing. I simply left a comment so future reader knew what that EECS acronym was (ie a mix of two different fields). – Andrea Lazzarotto Jan 28 '18 at 23:25
13

I suggest not worrying too much about it. Research is often never finished - you can reach a nice stopping point, but there's always something further you can try. It's not uncommon for a Masters student to graduate and then the supervisor engage a PhD student to keep developing the work.

If you can't complete the research in the remaining 3 months after three months of preparatory work, what are the odds that you'll be able to jump ship and complete a different project in 3 months?

Your supervisor should be aware of how much he can ask from a Masters student. If the conclusion is you've genuinely tried + learned something, but he gave you too ambitious a project, then you'll probably get the MS degree anyway. If you did better than expected of MS students, you might even get a MS degree with honours. Just relax, do your best, and don't worry too much.

  • 5
    Research is often never finished.” I think often isn’t necessary. – aeismail Jan 28 '18 at 17:08
12

tl;dr: Talk to your advisor about how to proceed

Indeed, as others suggest, yours is not an uncommon situation; and in fact many (most?) Masters' projects are actually 1.5-2 years long anyway. But, be that as it may - don't agonize over this alone: You have a supervisor. Use him/her. It's also their responsibility to make sure that you're on track, and it's also their fault if you've gotten stuck in a dead end. Tenured academics are also evaluated by the performance of their grad students, so there's also a "utilitarian" benefit to him/her helping you.

Your advisor is likely to have reasonable suggestiongs that are more specific to your situation rather than the wild guesses we're making.

3

Normally a Master’s by research is around 1.5 to 2 years long. Most meaningful research work comes out of this pure research program.

Yours is a 6 month research program as part of a larger coursework program. Thus I would not expect much in-roads into the topic (however there are a few people who really make nice breakthroughs).

The key is to know where to stop. It’s seems like you are working on a new thing. Thus merely summarizing and insignificantly expanding the knowledge of the current boundary seems to be fine! It’s an MSc not a PhD! :)

I know a friend who spent 4 years doing a 1 year MSc project as part of his 2 year MSc coursework degree. It really can get out of control. I don’t think he made a significant contribution even after all that time and I think he regrets it to this day. To top it off, he didn’t even have funding. Go figure.

  • Could you elaborate on the "normally [..] 1.5 to 2 years" bit? I suspect this varies per country as I've never heard of a MSc project exceeding 9 months on a 2 year degree, unless one studies part-time. – Gerard Jan 29 '18 at 10:54
  • Yes, a masters by research full time degree is usually around 1 to 2 years. It is basically a mini PhD, but if a student has got enough publications and if the results look promising, then it can be easily upgraded to a PhD. This is NOT a project. It is a full degree dedicated to research (coursework is optional or not needed at all). South Africa, India and Australia have these programs. I am sure many more countries offer this, I have only mentioned the countries that I’ve personally visited. Undoubtedly it may seem weird to American students :) – Harish Jan 29 '18 at 11:03
  • @Gerard The answerer is talking about the degree program, not the research project. – Stella Biderman Feb 13 '18 at 20:01
2

Talk with your supervisor about your concerns. It is his role to give you guidance on how to proceed, not just at the start but throughout the project. Communicate with him regularly, not just by email but especially in person. His job is not only to challenge, but also to help. I bet he's had his share of struggling students who reacted by undercommunicating for too long. (I have been on either side of this dynamic.)

Don't feel bad for appearing to be underperforming. Sometimes things turn out to be harder than expected – that's why it's called research. Show him why you think it can't be done within the 6-month timeframe, and what you reasonably expect to complete. It's his call to decide whether the problem is above or you are below expectations. In either case, keep him informed sooner than later, so he can make a better judgment.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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