34

I am currently writing my bachelor's thesis, and share advisors with a fellow student, who is also working in the same office as I am. We exchange daily about our progress.

I am under the impression that his work is going nowhere. He is working on a topic that has no previous work available, and according to him, direct verification on whether his research is useful is not possible. An indirect approach using yet another students' algorithm showed worse results in combination with the work of the student in question. To acquire this information, both students spent a large amount of time that could have been spent towards writing their theses.

He is somewhat backed up by his (our) supervisor, but I feel this is largely because the supervisor wants this research to end up in a peer-reviewed publication, even though there are no results that suggest the research is valid or useful in any way yet.

I have the impression that continuing this path will worsen the thesis outcome for my fellow student, as he is facing a deadline to turn in his thesis, but spends a lot of time trying things out and interpreting essentially bad results to show something positive.

Since I sympathize with him, I would like to intervene and talk bluntly about the issues written out above. I do understand, however, that it is not my place to correct him, as we are at the same stage of our careers, and I am about as inexperienced as he is. In addition, since our supervisor has a different opinion and more experience, it could very well be that I'll be wrong in the end.

What interaction, if any, is ethical and reasonable for me to engage in?

I imagine this sort of thing happens more often, even with more experienced academics, as confirmation bias seems to be a somewhat common issue.

How are similar problems usually handled in academia? Are people like my fellow student in question left alone to their own judgment, or do co-academics talk them out of seemingly bad ideas? Do people criticize each others' research a lot?

  • 19
    I do understand, however, that it is not my place to correct him, as we are at the same stage of our careers, and I am about as inexperienced as he is It is your place to correct him or, at least, discuss your concerns (which might be wrong), you don't need to be "above" someone to do that. – user2768 Jan 25 at 10:12
  • 5
    Offering unsolicited opinions about someone else's work is a violation of interpersonal boundaries. If you are asked for your opinion or guidance, then share your concerns honestly and openly but also tactfully. But, until you are asked for your opinion, it's not your place to share it. – Matt Jan 25 at 16:30
  • 25
    I do see a huge difference between stating your concerns and trying to correct someone! – cbeleites Jan 25 at 17:42
  • 23
    @Matt If I was working on a project and somebody else had a very good reason that my approach would fail, I would definitely want to know rather than waste years of my life. But it can be a delicate issue, so of course it should be approached with tact. – Jair Taylor Jan 25 at 19:30
  • 6
    "An indirect approach using yet another students' algorithm showed worse results in combination with the work of the student in question. To acquire this information, both students spent a large amount of time that could have been spent towards writing their theses." - that sounds like the perfect outcome for a research paper as well as the basis for a degree. Proving that a method doesn't work means that you've saved other researchers from making the same mistake – Valorum Jan 25 at 23:01
58

Which kind of interaction, if any, is ethical and reasonable for me to engage in?

Discuss your concerns with your peer.

How are similar problems usually handled in academia?

Through discussion.

Are people like my fellow student...left alone to their own judgement,

Largely, yes, but...

or do co-academics talk them out of seemingly bad ideas?

...in an ideal world colleagues (who they've discussed their work with) will try to bring them back on track.

Do people criticize each others' research a lot?

Yes! Constructive criticism is central to the research process (unconstructive criticism is unfortunately common).

46

It seems your main concern about your colleague's research is that the results are not "useful". I'm not sure what it means to you that the results aren't "useful" - is your colleague doing something that hasn't been tried before? If so, the work may be "useful" even if only to rule out a particular approach, or to clarify problems with a particular approach.

For example: suppose I have an idea to use X to address problem Y. It turns out that the state of the art, Z, is a much more efficient and closer to optimal solution to Y than X. However, by trying out X, I at least am able to 1) rule out the approach, and 2) gain some insight into why X does not work as well as Z for problem Y, what features of problem Y suggest it is not amenable to solution by X, etc. These are "useful" results (especially in the context of a bachelor's thesis).

Now, suppose I have a colleague who is in this situation, but who is still trying to somehow make X into a good (or better than Z) solution for Y. I would probably (in the course of normal discussions about our research) say something to my colleague like, "I think the most interesting part of your research is what it teaches us about problem Y, why X initially seemed like a good solution, and why Z turns out to be much better than X. If I were in your position, I would focus on that aspect as my main contribution, especially with a thesis deadline coming up." Often, people are so close to their research (whether their results are promising or not) that they may not realize what parts of it could be most interesting to the broader research community. This kind of feedback (from anyone - supervisor or colleague!) can be helpful.

  • 23
    The canonical example of research that didn't seem useful was work in optics in the 1950s, on highly coherent light sources. At the time, it was very much in the category of "Well, that's probably theoretically interesting but what's the point?" Now, we call it a laser. – David Richerby Jan 25 at 22:30
  • "It seems your main concern about your colleague's research is that the results are not 'useful.'" That is not at all the impression I get. The impression I get is that the asker's primary concern is that their colleague is torpedoing themselves by chasing a result that doesn't appear likely to pan out in the time available, maybe even to the point they won't be able to graduate as a result. The fact that there's no obvious application just means it's even harder to justifying continuing on that course. – jpmc26 Jan 28 at 12:38
  • Your follow up to that comment suggests that perhaps the best choice for the asker's colleague could potentially publish some kind of negative result. That does not seem to be what the colleague is doing, though. The colleague appears to be continuing to spend time looking for a positive result. Perhaps I've misunderstood, but if not, discussing that possibility with the colleague might be a reasonable answer for the asker here. – jpmc26 Jan 28 at 12:42
9

I would, gently, make your point known. Make it clear your point is in sympathy for him, not in directing him or arguing with him. Perhaps it will help the fellow. Perhaps not. If he doesn't want to listen, back off and just let the guy take his lesson. But I would probably go ahead and give him the heads up. At least he can consider the issue.

P.s. I think your instincts are good here. Project selection (reasonable scope, available resources, etc.) is a key to success in research. Just "doing what the advisor suggests" is not independent. You always have to decide if it is a good idea to work on a project someone suggests. Time is finite. Life is finite.

  • 2
    I'm in Germany (just as OP seems to be). In my experience, this level of independence would not be expected for a bachelor thesis (many supervisors and the time frame also wouldn't allow it). I'm going by the rule of thumb that the student has to do their own work, but a bachelor thesis is guided, a master thesis should not require guidance. Even for PhD theses, the leeway for the canditate selecting their project can be quite small in practice, depending on project funding/description (just as in real research life afterwards). – cbeleites Jan 25 at 19:23
  • Yes and no, on the guidance. There is a line between servile following and disregarding any input. I would argue you can/should have a perspective (just as an intelligent, adult, human) on the likelihood of success of your project. These things can be negotiated. You can modify scope, push back, decide a different advisor, etc. And in the end...it's your time. Gotta look out for yourself. – guest Jan 25 at 19:29
  • 4
    Sure. Maybe most of our different perception of the situation here is that I'm not yet sure how good OP's judgement of their neighbour's likelihood of success is. I did get the impression that OP misunderstands thesis success as positive or useful (?) results whereas the thesis committee is to judge whether the student did sound science. – cbeleites Jan 25 at 19:34
  • Oh...I had a German speaking relative who lived behind the Iron Curtain, for a few years. He said..."well, they can't vote in elections so they vote with their feet". Something to consider... – guest Jan 25 at 19:35
  • cbeleites, I just said he (OP) should bring up the issue. "At least he (student) can consider the issue." This is not an assertion of Euclidean surety. Capisce? – guest Jan 25 at 19:36
9

Your co-ed seems to be passionate about his research, and according to you it is original. His supervisor seems to be satisfied with his work, so far. Positive results are not neccessary for a bachelor's thesis, and results are not usually expected to be published. I don't see a major problem here. Certainly no reason to "intervene"! So if you are interested in his work, ask about it, have him explain it to you, maybe bring up ideas, offer to proof-read whatever he has already. But don't add any pressure.

  • 4
    Exactly. Bachelor theses are a great way to do a proof of concept for high risk ideas and it's no problem if it doesn't result in anything more than the student learning something and knowing that the project doesn't work in the way envisioned. – DSVA Jan 25 at 21:04
8

Not getting nice (positive) results during a thesis is, well, not as nice as getting positive results.

It often also means more work for the student: after all, no nice results could be due to either

  • the previously unknown reality just really not being that way*, or
  • the student not being up to their task.

As the thesis is part of an exam, the student has (or feels pressured) to make sure the lack of nice results is not ascribed to their inability. Hence the additional work to make sure no examiner gets the wrong idea.

But it is perfectly possible to demonstrate sound scientific working (which after all is the exam task) while showing how something does not work. There is IMHO nothing inherently scientifically bad in their thesis as you describe it.

BTW, I'm speaking as someone who got a perfectly fine Diplom on a thesis showing how the approach suggested by my supervisor did not work out. By now I can even tell, why ;-) - and it triggered the better part of my PhD thesis and is still a pet rearch topic of mine years later.

Bottomline is that in this part of your description I do not see anything that clearly indicates** a need for "correction".
There may be a need for encouragement and understanding that that thesis may be even more stressing than other theses are, though.


That being said, what kind of raises a standard red flag is the approaching thesis deadline. Most students I've seen got in trouble with this deadline and the writing up of the thesis.

Maybe a buddy system in your office could help? As you already update each other regularly, maybe you could bring up the idea whether it would be good for both of you to remind each other of actually writing your theses? (Regardless of results, e.g. theory can usually be written up rather early.)


* It was quite an eye-opener to me when someone from a funding agency once told me that they would fund projects only if they estimated the probability of failure to achieve the proposed [whatever] at least 80 %. At the same time, I'd have said that the inner-academic view of just such projects was that they almost always succeed...

** Clearly as in I'm sure the problem is with your fellow student's thesis and not triggered by a misunderstanding of what a thesis should be and what the odds of doing sience are. This is partially triggered by your writing

  • "interpreting essentially bad results": the results may be negative or not be what is desired, but the only way they can be bad here is if they are wrong because of bad science.
  • "there are no results that suggest the research is valid or useful in any way yet". Usefulness is not a useful criterion to judge the value of the scientific work here. Results can of course be valid or invalid, but it's not the results that can suggest they are themselves valid (by being positive or nice) - validity has to be established independently of the outcome. The important question for validity and the exam is: does the student do sound science or not?

Of course, this could also be a bad thesis, just like a thesis with nice and "useful" results that is scientifically unsound.

  • I think the person from the funding agency who told you (*) was pulling your leg. – user21820 Jan 27 at 2:33
  • @user21820: I don't think so. They certainly expect that there will be proper research, so results and scientific knowledge gained. Including a bunch of theses. BTW: This was funding for applied projects, with rather specific final milestone. The idea here is trying to bridge the "valley of death" between reseach not being basic any more, but not yet sufficiently advanced so that industrial research can take over. So, also, tax money would be wasted on projects where the probability of success is sufficient so that industrial research can take care of it, which corresponds to the probability – cbeleites Jan 28 at 8:52
  • I remain very skeptical. As you wrote it, since 100% > 80%, if they believed that a project was absolutely certain to fail, they would fund it. And if they believed that a project was certain to succeed, they would refuse to fund it. That's ridiculous to me. – user21820 Jan 28 at 8:55
  • ... of success (or failure) threshold. The actual threshold may very well be field-specific. Antoher reason to not have the threshold too "easy" is that funding projects with high success probability would be a more-or-less hidden subvention of the involved industry partners which would be illegal due to market distortion. BTW, I did not say they'd fund every project they think will fail. E.g., they'd certainly refer projects they consider not sufficiently advanced back to basic research funding programs. If the project were certain to succeed, the industry partner(s) would do the R&D ... – cbeleites Jan 28 at 9:14
  • ... on their own. After all, they can just hire scientists or fund a PhD or pay a Fraunhofer institute to research for them. That is done all the time, and sufficiently certain may very well be 1 out of 3 or 4 such projects. – cbeleites Jan 28 at 9:17
6

Other answers have done a pretty good job of explaining that everything you've described is actually valid and useful research (and you've said nothing to suggest that confirmation bias is actually a factor). So I just want to point out another problem you seem to be overlooking in your idea to "step in".

The problem is that if your fellow student abandons his project at this point, how is he supposed to complete a bachelor's thesis at all? At this point in the year, I guess he's used up significantly more than half of the time he has to actually do the research for this thesis. Starting a new one now would not just waste the valid work he's done so far, it would also leave him in a terrible position to complete a different thesis. That result would surely be worse than this one.

So step back and consider what the real objectives are here. Ultimately, the bachelor's thesis is not expected to change the world. There are two primary objectives: (1) to give the student an opportunity to try his hand at open-ended research where the answer is not already known; and (2) to give the advisor a chance to see how the student does given the project he's working on. This project is all that is needed to achieve those objectives. Trying to change horses midstream would only confuse things, and would not be helpful to those objectives.

By all means, discuss his research with your friend, try to understand it for what it is, and try to come up with ideas to make the research even more useful. But don't suggest that it's not useful. You wouldn't be "correcting" him by saying that; you would be misleading him. And certainly don't try to find ways to spin it so that it sounds good; the research will be valid and useful if and only if he is honest about everything he has found. Understanding why this algorithm isn't as good as some other one could lead to deeper understanding and more insight to tweak one of the algorithms so that it's even better — if not by him, then possibly by others who read his results. Or at the very least, his research could let others who are wondering about this approach avoid wasting their time. That's entirely valid and useful.

2

You shouldn't interfere. Getting trapped into rabbit holes is part of the research process. Learning how to get of one on your own is part of the experience. It's a bit like trying learning to walk without falling; or walking only on a perfect surface.

Sure, someone could hold your hand all the time; point you the obstacles on your way etc. The point of research is to gain autonomy in your thought process.

Your supervisor probably knows what's going on but won't step until it's absolutely necessary. If you are asked for help, do share your views.

In academia this is handled by the researcher requesting feedback from his/her supervisor, from his peers, presenting at workshops and conferences etc.

For now your best help would be just to listen to him, and ask questions about his research without judgment of the results.

  • 1
    This seems like an awfully wasteful stance... – user2768 Jan 25 at 12:55
  • 4
    @user2768 I argue that going down rabbit holes is useful, you may discover useful things. A lot of useful research has come out of rabbit holes. Non-euclidian geometry comes to mind. – Koenig Lear Jan 25 at 15:03
  • 1
    I don't think it's wasteful. Rabbit holes can be an important part of the learning experience. Allow the adviser to be the guide. Don't assume just because you don't see value that there isn't any. – scrappedcola Jan 25 at 15:03
  • 1
    @scrappedcola (and Koenig Lear) I'm not assuming because you don't see value that there isn't any, I'm suggesting that if the OP doesn't see value, then they should raise their concerns. Following the disclosure, the researcher might discover whether they are going down a rabbit hole, which has significant value. – user2768 Jan 25 at 15:55
  • 4
    @user2768 Don't forget one basic fact: if you know in advance how what you are doing will turn out, then you are not doing research. – alephzero Jan 25 at 19:52

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.