My supervisor gave me some guidance in order to complete my Masters thesis and to publish the result in a journal. His idea to develop the work is not correct and in this case I know my thesis better than him.

How do I tell him his idea is not correct?

  • 45
    No one can answer this without knowing the culture of your field, the culture of the country you are studying in, the personality of your advisor, the nature of your relationship, and so forth.
    – Corvus
    Commented Mar 29, 2015 at 4:15
  • 17
    Your argument "I know better because it is my thesis" is not valid. Can you actually prove why his idea is wrong? If yes, show him your unquestionable proof. If not, you would probably have to implement his idea and then we will see who is right and who is wrong.
    – Alexandros
    Commented Mar 29, 2015 at 15:03
  • 5
    @Alexandros Your reading is not valid. The OP never said that he knew better because it was his thesis. Rather, the OP explicitly wrote his ideas to develop the work are not correct and that he knows his thesis better than him. He is not stating the latter implies the former.
    – JNS
    Commented Mar 29, 2015 at 16:45
  • 13
    By giving proof to that effect (assuming you're in anything at least semi-formal) and going about it tactfully? I.e. don't go "you're wrong, see how I debunked it?", go "it looks like this approach is flawed, I found this counterexample and I don't see how we could fix that".
    – G. Bach
    Commented Mar 29, 2015 at 19:41
  • 8
    In the egomaniac case, the "I'm not sure I understood your approach correctly, I can't figure out how to make it work on this example" version may work better. Commented Mar 30, 2015 at 11:37

9 Answers 9


I had some serious qualms about this when I started out research in undergrad. My advisor had some pretty mistaken notions about the topic we were pursuing at the time. He is a very approachable and understanding guy but I was still a bit hesitant to say anything. After all, he was the established expert and I was the lowly undergrad. How I approached it was along the lines of: "I'm not sure if I'm doing this correctly but would you mind checking my work? Maybe you can explain what I'm doing wrong."

You don't want to challenge or outright claim your advisor is wrong. You wouldn't do this to a friend¹ and you really shouldn't do it to an advisor - it's a good way to ruin your professional relationship. If you're right, you've come off as very arrogant and create some animosity between you and your advisor; if you're wrong, you've let your arrogance get the better of you and killed your credibility.

Professors are human and they can be mistaken more often than you'd think. You're going to run into this a lot more than you might have thought. After a while, you can start to be more upfront and honest after you've established plenty of rapport, but early on it is a bad idea.

¹ I've done this and nearly ended a friendship with a very good friend because my arrogance got the better of me.

  • 36
    I disagree with this. As long as you keep a discussion on the scientific level, no one should get offended. Of course, one has to keep it polite.
    – Davidmh
    Commented Mar 29, 2015 at 8:41
  • 2
    @Davidmh Especially because this is work. We're not pursuing friendship - we're just stating the facts in order to get the job done. While I do disagree with the post I do see it as a good way to approach your superiors if they happen to be egotistical. Commented Mar 29, 2015 at 15:25
  • 22
    @Davidmh I agree no one should get offended but from others' experiences, I think this is the best course of action. Some professors do not take criticism from their students very well. It's not very common but it is better to be safe than sorry. Commented Mar 29, 2015 at 15:29
  • 5
    @Davidmh - Maybe you were lucky with your advisor, but not everyone is magnanimous enough in the scientific world. In general, pseudo-scientists have monster egos, and they are a part of Academia. Not everyone responds positively when faced with their limitations/errors, though there is no real sin associated with making a mistake. The question pertains precisely to these kind of people IMO.
    – 299792458
    Commented Mar 30, 2015 at 6:45
  • 3
    @Davidmh One shouldn't assume that the advisor is an egoistic monster, but given that such creatures exist, it is reasonable for one to take precautions that the advisor might be an egoistic monster.
    – KSmarts
    Commented Mar 30, 2015 at 17:27

The safest approach is always to make it a question about you rather than about them. "I'm confused. You said... but I had previously read/been told/seen ... instead. The two seem to conflict; can you help me understand what I'm missing, or what that other reference was missing?"

They may be able to show that there isn't actually a conflict. They may be be able to show that you misunderstood something, or that the other reference was outdated, incomplete, doesn't apply in this situation, is an alternative theory that they disagree with because.., ... Or they may say "oops, you're right."

By not accusing them, you avoid embarrassing yourself if they were right, give them space to correct themselves gracefully if they were wrong, and maximize your own opportunities to learn.

  • 1
    This is the obvious answer in my opinion. The student should assume themselves likely wrong, and thus approach the contradiction in an inquisitive, not accusatory, way. The only exception here is if the supervisor can be logically proven wrong, but even then, it's well possible to write false proofs. As such, even when one sits there with what can only be understood (by the student) as a proof that they're correct, thus making them more likely than not to be correct, they should still ask, and not accuse.
    – user110391
    Commented Feb 17, 2022 at 20:37

The answer can only be politely, respectfully, humbly, and open to the idea that it is you, in fact, that is wrong. There might be something you don't understand and which your advisor can correct. Or maybe your advisor really is wrong. These kinds of academic conversations can be difficult but they are part of the most exciting parts about doing science and some of the best opportunities to learn.

  • 1
    Excellent, brief and to the point.
    – Floris
    Commented Mar 30, 2015 at 16:42
  • Who knows whether the OP is right or wrong? Certainly not @Benjamin Mako Hill (he can't be a clairvoyant). The professor is probably right, and I can understand that. But it is ultimately up to the OP to decide rightness for him/herself. Alas, we are collectively in a culture that gives too much importance to the holders of knowledge, and not to knowledge itself. The kind of attitude expressed in this post makes me sorry for the egocentrism of Benjamin and his ilk. But then, I care only for the student OP.
    – PKG
    Commented Jan 29, 2016 at 1:53
  • 1
    @PKG: I said be open to the idea you are wrong. I didn't say that OP was wrong. I also said that maybe your advisor is wrong. Of course I have idea and I never said otherwise.
    – mako
    Commented Jan 29, 2016 at 7:20

At the end of the day, science is not about people. The universe is what it is, regardless of what we think of it, and science is simply our means of discovering these facts.

If your supervisor really is wrong, this must result in something which has evidence that can be observed. Moreover, you must have seen evidence that convinces you to have a different opinion from your supervisor. Leave the personalities out, and get at the evidence. If your supervisor is actually wrong there must be evidence that can be examined clearly and impersonally. If you are merely convinced to place your bets differently than your supervisor, then it is only a difference of opinion until the evidence can be gathered.


I had a lot of such discussions with my supervisor. He knows the field much better than me, so I never thought he was mistaken on specific facts. Instead, most of our disagreements were on what the next steps would be.

S: Maybe you can improve the second step using [algorithm].

D: No, that is not going to work because [...the data is too complex...].

This usually started an engaging discussion on the details of the data, alternative algorithms, and ways to circumvent it altogether. For this to work, you have to listen carefully to what he says, have an open mind and learning attitude, be polite, and of course, keep the discussion on a purely scientific level. Your arguments should be supported by evidence, and it should be clear how strongly supported they are.

Also, I think you should indicate how convinced you are when you phrase it. This way you give a true idea of how confident you are, and how much thought should they put into your idea. Furthermore, if you confidently assert something that is factually wrong (and this doesn't happen too often), you get a better chance of getting a detailed and enlightening lecture on why it is wrong.


Discuss your concerns with him obviously, in a courteous way, and in person. There are two possible outcomes: either you are right and they'll be grateful to learn early on; or there is potential that you don't see, but your adviser correctly identifies. My hunch is that if your adviser genuinely thinks that what you are working on merits publication once fully fleshed out, they are probably right.


If you understand the core of the problem well enough, and discussions with the supervisor were not successful, then you should take your time to write down the argument you want to make in pretty much the same style as you would write down a scientific article. You then ask your supervisor to consider what you have written down, he'll then have to read this when he has the time for that.

It may be that he'll try to do that during a meeting, or he may take quick look and then decide that he has to study it in detail and come back with his thoughts later. Either way, he is then forced to read what you wrote and then ask you to clarify it point by point wherever he has objections to the way your argument proceeds. What matters is then that the conversation will be driven by your line of thoughts and not by his ideas. In a typical supervising session this may well be the other way around, preventing him from getting your point.


Considering this is your thesis work and that the supervisor is counted as an asset(in most cases), do try and let him/her know about your ideas. Do question the ideas on why he/she has mentioned their ideas and why your idea could be more suited to this particular application.

If it were me, I would do this:

"Sir/madam, you had mentioned these points on how to go about the thesis and had given me some great ideas, thanks! However, could you please tell me if this approach of mine would hold good in this application? I did draw a parallel between both our methods of implementation and I felt approach X could be done in this way since (reason1,reason2 etc). Please give me your opinion on this since it will improve my understanding a lot more.

PS: I am personally of the idea that one must not come off as too arrogant or revert back with off the cuff remarks considering this is something I consider very professional in nature. So there is no proving one another wrong, rather, work together to reach a better understanding. As someone said earlier, no matter if one is an expert or a student, we are all work in progress.


I would like to uplift answers above that advise the gentle It's not you its me approach. I've done both the I'm right you're wrong and the former approach. Both have worked. But when I'm wrong with the latter approach I end up being more embarrassed, humiliated, and anxious about continuing my work. It has left me feeling that my advisors trust my work less and are less open to me pointing out errors.

In the end science IS about about people: egos, personalities, personal beliefs, and competition whether implicit or explicit to the process. Understanding how to navigate these social biases and situations per person, and at a systemic level, is important for not only being able to work well with other people but more importantly achieving the poster's original goal: furthering their MS work and their science they are researching.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .