63

I got my PhD in data engineering in 2014 from a reputed university in a developed country. I am now an associate professor in a well-reputed university. Between my PhD and my current position, I took several postdoc positions in very good universities in several countries.

The papers with which I got my PhD (with distinction) are problematic. These papers were published in very good journals and proceedings. However, I didn't conduct the experiments as I mentioned in the papers. They aren't fully falsified, I had assistance and some are true only for specific parameters. I knew this when I published.

More than two years, I began to feel terrible about this. I feel that my PhD is fake and everything I am doing now is based on academic misconduct.

I now excessively try to make everything very correct in my papers and to be extremely honest in presenting the results. I often disagree with my coauthors when they want to publish a paper with weak experimental validation. They always argue that at this rate, we can publish only a few papers and we can never compete with our fellows.

I don't think this will ever be discovered. Even if you execute the experiments, you always find different results, the details of the approach are not given in the papers so any missing details can change the results.

What should I do now?

5
  • 14
    The title says "many years later" but your fourth paragraph seems to suggest it was only two years ago. Am I misreading this?
    – Michael
    Jul 7 at 22:26
  • 2
    @Michael "Now, I feel terrible about these papers [...] for more than two years already and I am wondering what I should do. " The pain started two years ago.
    – EarlGrey
    Jul 8 at 13:37
  • I find it strange that most of the answers you receive focus on the emotional/evaluative response of guilt (not a word you actually used) rather than focusing on the final sentence ... "what should I do now", which I take to mean, "What should I do now,about the fact of there being in existence papers whose results are incorrect or false?"
    – user02814
    Jul 9 at 3:43
  • How hard is it for a reader (who, say, is doing a thorough literature review on the subject) to arrive at the conclusion that your papers are crap?
    – Karl
    Jul 9 at 10:49
  • Btw. what does your old mentor say about the subject? He's the one who should be ashamed.
    – Karl
    Jul 9 at 11:06

13 Answers 13

66

Separate two things: The guilt over having obtained your PhD in a dishonest way, and the guilt over having falsified papers in the literature.

First, the PhD. While you may not have deserved your PhD at the time you received it, it seems that you have accomplished good work afterwards. You are qualified for your role now, and there isn't really anything you can do to fix the past. If you need to atone, consider making a sizable (compared to your disposable income) donation to charity - maybe something helping disadvantaged students to access university education (in case you feel guilt over having "taken" a spot from someone else)?

Second, the papers. This is a different situation, because the continued presence of falsified results in the literature is an ongoing harm. You could perform the experiments now, and issue a correction with the actual results. If you have omitted any restrictions from the paper, issue a correction stating those. If it is not possible to fix a paper, you need to withdraw it. These actions obviously come with risks to your reputation, but there is no way around them if you want to do the right thing now.

10
  • 38
    I think there is another aspect you might want to mention: do others also perceive that these papers contain misconduct. Maybe they follow the standards of the field but Alice's personal standards are higher.
    – Christian
    Jul 6 at 10:29
  • 9
    @Christian The OP is clear about having published partially falsified data. Whether or not their immediate subfield considers this aceptable is irrelevant.
    – Arno
    Jul 6 at 14:54
  • 19
    The OP also wrote "After that, I got a phobia and I was excessively trying to make everything very correct " - this is what made me take the previous statements with a grain of salt.
    – Christian
    Jul 6 at 15:12
  • 21
    Wait, wait, "If you're feeling guilty, just buy off your guilt with the money you made from the very thing you're guilty about" doesn't seem like very ethical advice.
    – corsiKa
    Jul 7 at 3:06
  • 19
    Once you get beyond the standard knee-jerk reactions here, welcome to the beginnings of the actual human experience. You made a mistake, bless you, turns out that’s the stuff that makes you, and the rest of us, less than the angels. Don’t burn in Hell, just learn. Jul 7 at 4:51
17

Your guilt is rooted in the fact that you completed a PhD without taking the time to properly (according to your standards, standards that are higher than your colleagues and mentors at the time) complete the papers.

So you feel your current professorship is undeserved.

Ask yourself: if at the time you had enough time and resources, would you publish proper (according to your standards) papers? From what you describe, I would say yes. And this is the most important teaching you have to share and enable with the M.Sc and PhD students you are tutoring, this is what makes you valuable in the academia. You learned it the hard (wrong?) way, by doing it, make this step unnecessary for others.

Give your students resources, give them time, take into account for each PhD a buffer of 6 months to wrap up (additional to whatever length is average in your country) read the papers from candidates to the positions you have open, do not simply read their h-index/citations or other metrics that are there just to be gamed (and the way you published during your PhD is the consequence of these metrics ... it would be extremely irrational not to game them).

Regarding doing science, you are evidently on par with the others, since you obtained various postdocs and you landed a professorship, so you did not take a valuable spot.

Final note:

if you execute the experiments, you always find different results and the details of the approach are not given in the papers

So your papers are very weak, or even inconclusive: these papers are needed in science, because they can trigger discussion. Is this a waste of money? in the world where results are the goals of science, yes, and if you feel like that you would find yourself better aligned working in the industry, in a world where science is the goal of research results, one cannot expect to have high impact results from all the funded research...

13
  • Why would it be "extremely irrational" not to game the h-index/citation metrics? Yes it can be done, and yes it isn't hard and yes those metrics are bad for a multitude of reasons, but that doesn't meant the "rational" course of action is to game them! Is that really what you meant to say?
    – terdon
    Jul 7 at 10:42
  • 1
    @terdon what else is rational, if your career depends on some stupid metrics and you are at the bottom of the scale, with no power of changing it? Remember, we are discussing about PhDs and the like, they are already slaves of lousy working conditions, although they have some marginal control on that, expecting they immolate their career to prove those metrics are useless is not something I force on them. From higher rank academics, I absolutely expect them to fight against these metrics. And they don't. For accepted rational reasons like:
    – EarlGrey
    Jul 7 at 11:27
  • 3
    I am not defending the metrics. However, gaming them is immoral and I see no reason to call it "rational", sorry. Falsifying information and unethical conduct is very hard to defend as "rational". I'm sure some people do it, but that's their problem and we shouldn't promote the practice. (Also, my name is terdon, not tendon :) ).
    – terdon
    Jul 7 at 11:59
  • 2
    @terdon: I think there is a difference between "making sure that a crappy paper is published in a prestigious journal to get the credit * index" and "falsify the results". The first one is a normal consequence of the idiotic way science is measured and everyone will make attempts to get whatever published. It does not mean falsifying the results, it may just mean making extraordinary claims and hope for the best, or publish nice but useless information. This is the sad reality of a failed system.
    – WoJ
    Jul 7 at 16:28
  • 1
    -1 for the cynical answer that says in essence "the careerist ends justify the means" and "bad or even falsified research might be useful, don't bother correcting it". Jul 7 at 19:23
15

Support reproducibility

First, I subscribe to @Arno's answer and I have already upvoted it. I am also in data science at a reputable Western university and what he/she advises are the standard practices in this field. I would just add that, if you find that the journals where you have published your papers do not accept corrections, you can post the corrections on your web page, along with links to the original papers.

Moreover, you may want to learn about the reproducibility movement and promote it more systematically with your collaborators or in the classes you teach. For instance, I teach a project class about research practices; in it we also discuss publishing reproducible research. Advocating for reproducible research will not fix your previous papers, but will improve your field of research as a whole, and will make it less likely that the misconduct you committed is accepted (see e.g. Donoho's paper below).

One classic article I use is An invitation to reproducible computational research by David L. Donoho. In machine learning, from 2019 on, conferences have introduced reproducibility checklists and reproducibility challenges, as shown in this paper. For a behavioral research themed discussion you can look through the posts of datacolada.org. I am citing the last source because psychology was one of the first fields where the reproducibility crisis was openly recognised, and you are more likely to find concrete suggestions.

Hope this helps. Science makes mistakes, but it corrects itself. Humans make mistakes, but they can learn to become better. [Please allow for the following less than objective insertion: I respect you very much for caring enough to do something about research ethics.]

I also suggest that, if you aren't doing it already, you take on peer reviewing as a service to the community. At the serious journals and peer reviewed conferences, your work ethics will be valued.

6

obtained with some assistance and some of them are true only for some specific parameters

the details of the approach are not given in the papers so any missing details can change the results

If these are the worst parts of the papers, don't worry too much. It is a widespread problem in science that cherry-picking of "statistically significant" results occurs, and that reproducing results is difficult.

And indeed it is a big and important problem, but you are already on a good path by recognizing the shortcomings of your prior work.

I would suggest that you check if your prior publications get referenced in other work. If not, don't worry - the results apparently weren't very important, and the main result was your personal learning.

If they do get referenced, read the new publications to see if the shortcomings in your work could affect them. If it does, then you should work on making a follow-up that will point out circumstances under which the original research is valid, and also that will make it possible to more easily reproduce the results.

6

Cutting to the chase here, you cheated and you obtained very large benefits from your cheating. The first question, surprisingly unexplained by you, is exactly why you did this. Please say something about this.

Cheating carries unsuspected penalties. Building a career in academia is grueling and exhausting. People who jump through all the hoops successfully end up with increased strength of character, determination, and persistence. In fact they change remarkably during those long years of training.

When you cut corners or cheat to get ahead, you eliminate yourself from the group of people who came about their credentials the hard way and built the skills and character needed for a successful career. The demands don't stop with the PhD; for many people things get harder after that when they're faced with the larger demands of tenure, promotions, and more sophisticated research and publications. The competition continues too, such that those who aren't willing to put in the work or who don't have superior skills (because they didn't do their own academic work earlier) are culled from the pack. When things get really tough, those who learned honest toughness at the lower levels of their training pull ahead of the rest.

As you've found out, the psychological burden of cheating is large. There's small guilt and Big Guilt. How big is the fault in your case? Well, it's enough to get you ejected from your career and to have your PhD revoked. I'm really wondering why these thoughts didn't cross your mind when you were contemplating cheating. Please say more about this.

What to do now is a moral and philosophical question about your basic beliefs about life. For that reason, I suggest that you post your question in a moral and ethics forum or, better, get guidance from a skilled person like a therapist, life coach, minister, etc. I think you should start by asking yourself, "Why did I do this?" What does your conduct say about you? Are you happy with that? Next I would assess whether actual harm was done to others and the extent of the harm. Then, is it possible to make reparations to those who were harmed? Finally, ask how you can make this right within yourself. Many people have committed serious errors. They must find a way to make peace with their actions and move on. You will never forget this error you've made, but perhaps you can come to terms with it by correcting whatever flaw within you allowed it to happen and by accumulating enough positive actions to offset it.

At the extreme end, the hardest question for you is whether you would be willing to set everything back to moral zero by admitting your mistake to your PhD institution and accepting the consequences, which I think would be grave. This is a complex moral question that can only be answered with considerable introspection.

1
  • 2
    "When you cut corners or cheat to get ahead, you eliminate yourself from the group of people who came about their credentials the hard way and built the skills and character needed for a successful career." xkcd.com/1827
    – EarlGrey
    Jul 7 at 6:45
5

A lot of answers here are suggesting a middle of the road fix - donate to charity, see that you've done good work since etc. quite frankly this will (probably) not work.

If you made the mistake of thinking the papers were of sufficient and publish worthy quality, not knowing of any errors then you could quite easily fix the issue by issuing corrections, withdrawals etc. but for this is not the case.

The issue here is guilt. You feel guilt for your intentional wrong doing. Giving any amount of your income to charity won't help with the guilt, because it doesn't rectify it. Here's what could:

  1. Redo the bad work in full, at your own expense
  2. Go back to your issuing body and come clean. Explain if there were any circumstances that lead to your poor judgement, but own up. Tell them you have since fully corrected the work and are ready to issue corrections etc.
  3. Follow the advice/suggestions/enforcement they tell you to

This will mean you've not only paid for tour wrong doing, you've made good on your previous commitments.

5

The way you have framed this question, you make the "black and white" answer seem to be: come clean and announce to the world publicly that you committed some kind of misconduct with the papers you wrote and issue a correction.

However there are several factors here that I am reading between the lines that make me doubt that the "black and white" answer really applies here.

  1. Guilt -- especially guilt over something that has been festering -- can skew your perspective on how important something is. It's not clear to me, in objective terms, how serious the ommission is, but my suspicion is that it is fairly minor. To me, "major" would mean that the main result of your paper is not reproducible even in principle; "minor" would mean that you did not provide enough information to fully reproduce the results but that the main conclusions are correct and reproducible (perhaps with some additional input that can't be reconstructed from the text of the paper).

  2. Our own faults are very clear to us, while the faults of others are often less obvious. Are you sure that your assessment of your work is fair, in the sense that other papers in the literature do not contain similar omissions/simplifications? This is particularly the case since according to your description, your standards are higher than many of your peers.

  3. Science can be brutal, and we are hard on ourselves. People make mistakes, and deserve the room to learn from those mistakes. Be kind to yourself.

Furthermore, it's important to be accurate in your assessment of this, because "going too far" in the direction of "apparent good" can lead to damage to your reputation -- if you tell people you are guilty of something, people will tend to take you at your word. Admitting to falsifying results in a paper, even approximately falsifying them, is a pretty major offense in the scientific world.

I think you need to evaluate as honestly as possible, without guilt, whether the papers currently in the literature actively causing harm. Are people citing these papers? Are they building experiments based on the parts you know are wrong? And... realistically, do you believe your omissions are significantly larger than others in the literature in your field? If not, then let it go. I promise you, you are not the only one who has done something like this, and from your description I seriously doubt that what you did is the worst thing that exists in the scientific literature.

If there are really wrong results in those papers and those wrong results are being used as a basis for follow-up studies today, then that does make the situation more complicated. But I would still proceed cautiously. I would publish an erratum or some updated paper addressing the issue, but I would strongly recommend to keep the focus on the science, and not admit fault or falsification of the results. Just state that you became aware of errors or omissions and you are submitting a clarification.

The problem you have is that by raising the issue now, particularly if you frame the issue as something serious like "falsification", years and years after the paper was published, is that (a) people will wonder why you are brining it up (they may question your motives), and (b) you bring attention to the issue. If harm is not actively being caused, and your guilt is causing you to overestimate how significant the actual offense was, then there is a serious risk that the net result will be self-sabatoge without helping yourself, anyone in community, or, I would argue, even "the cause of justice" in an abstract sense.

You seem to be a very serious and dedicated scientist. Focus on doing good science and bringing positive and exciting new results into the world. Don't let this consume you.

2

I don't know what you can or should do here as regards how you feel about things.

But one thing. Do not become the type of professor who drives other people nuts for "not doing things right".

  1. It won't get to the heart of your own guilt - it will just enlarge it.

  2. It will cause other people in your department to wonder about you, why you are the way you are and your background - personal and professional. This could lead them to suspect what you already know about your doctoral work and its shortcomings. In academia, shared suspicion can be as damning as proven guilt.

On a human level, have you discussed this with your husband/wife or any close academic friend ?

1

Reading your post again, I see that your position is that you've reformed yourself and you're now doing honest work. What concerns me about your reform is that it was done out of fear---"I got a phobia." Though you say that your misconduct (your word) can't be discovered, it appears that your fear is based on the risk that it will be discovered and that you will suffer the consequences. You ask to not be judged. It's hard to withhold judgment when you haven't expressed any specific moral regret about your actions.

Data engineers deal with things that directly impact public safety, like self-driven cars. Now imagine that your surgeon, who's about to repair your heart valve, got his medical degree the same way you got your PhD, by not doing proper research or falsifying his studies. Hence he didn't really get the training he needs to operate on you. This is a problem, no? If he teaches in a medical school that's a bigger problem.

Since you're convinced that your misconduct can't be detected, what exactly is the problem in your eyes? You seem to be looking for exoneration. What exactly would justify exoneration in your case? Is there some special circumstance that would exempt you from the usual penalties in cases of academic misconduct?

2
  • 4
    Your answer works within a framework of retributive justice, rather than restorative justice. I think that the OP should work to repair the harm they have done (as mentioned by many other answers), rather than "pay the price" for what they have done. Jul 6 at 21:26
  • 1
    I actually agree that restorative justice is a better Plan A, and my post below (which was actually my first post) reflects this in part. I'm not sure it's possible here because the damage done can't really be calculated. I also feel that restorative justice needs to be undertaken soberly, beginning with a self-examination of the causes and impacts of the deed. I don't see that in the original post. Compare, for example, to racial restorative justice: reparations payments will help, but they're not enough without heartfelt acknowledgement of the harm done by oppression.
    – Eggy
    Jul 7 at 14:11
1

Let's break down the aspects:

  • You (co-)authored a paper which contains scientific misconduct, which you conducted
  • You co-authored other papers which did not meet your standards
  • You authored other publications which you PHD is based on correctly
  • I assume that, even after removing the tainted contribution, your PHD holds enough merit to justify it

Let's see what you could do:

  • Ask the ombudsperson of the institution in question
  • Inform the editor of the tainted paper on the nature of the misconduct and ask for retraction if you believe that the core content doesn't hold
  • Inform the editor about the fact "that the results are not reproducible" and suggest a correction
  • Write a comment (if such a mechanism exists in the publications in question)

However,

  • you should try to do this via the corresponding author
  • I read between the lines that the ethical standards of you co-authors may be not that high, so be prepared that they will block it or, if you circumvent this by going to the editor directly, try to throw you under the bus, which is more problematic if your PHD supervisor is directly involved (they may try to attack your PHD).

When it comes to your PHD, it depends a little but how much of it was tainted and if it was hard misconduct (falsification, fraud, plagiarism) or soft misconduct (cherry picking as far as usual or expected in the discipline). Imagine throwing away all parts of your PHD thesis affected by potential retractions and see if it still has enough value.

If you are independent enough (e.g. postdoc in different field), most of your PHD is not affected by the really problematic parts you can ignore the personal feeling of your former workgroup, but that also depends much on the location of the university.

1
  • The OP's wording leads me to think this is a case of hard misconduct. First, it's not one paper, but "papers." Second, he didn't "fully falsify" them; therefore he "partially falsified them."
    – Eggy
    Jul 8 at 9:33
1

they were obtained with some assistance

That is not bad. Scientific research is usually conducted in groups. If you are feeling guilty about not crediting someone you got help from, consider that if they were relevant enough as contributors, they would have been included as authors (PhD students, in my experience, are not the ones who choose that).

If they were not, it is true that you could have included them in the "acknowledgements" section (if that is used in your field), but that is mostly ornamental when you are thanking people, so you shouldn't worry. You could always thank them, probably even now.

and some of them are true only for some specific parameters

That is not fabrication. If you didn't provide enough information to reproduce your results, it just means that people may refute you some day. If you did provide the necessary conditions, it makes those papers more specific (and less interesting), but they are still valid.

What you seem to be worried is that you knew about these limitations, and decided to "hide" them, and feel that you were dishonest. There's always the option to publish a new paper continuing your research on that topic, explaining how those elements affect your research -be as exhaustive as you need this time!-. That would certainly remove the risk of someone extrapolating your results (although again, if they did, they would just find out that what you published is not as general as stated).

1

Exploit your unique* experience to better the world

* Ok academic misconduct is definitely not unique but go with me here

I've seen people who change careers, or people who have done regrettable things, do things using their past experience that nobody else would have thought of. Whether it's a dentist that becomes a software developer and solves some major problems in his industry, or a murderer that becomes a lawyer or journalist and fights for a better system, these people bring a unique insight that allows them to do things most people can't.

You would be better placed than most people to answer questions like:

  • how widespread is the problem in your field
  • what impact does it have
  • what factors contribute to the psychology of academic misconduct
  • what can be done to change the culture
  • what can be done to minimise the harm/detect occurrences
  • what is an acceptable level of rigour (i.e. what is the sweet spot between invalid results and slowing down valid scientific progress in the field)

Trying to impose standards on your collaborators seems a bit of a waste of potential. If the system is generating strong pressure to fabricate results, people within the system should be having open discussions about how to improve the system. You could be one of the people driving these conversations across your field as a whole.

To go public or not?

Of course, going public with what you have done puts your career at risk. It is not something to do lightly. But remember, how you frame things can make a big difference to how they are received. If you tell people you fraudulently obtained your PhD, you will likely be in a lot more trouble than if you tell people you cut some corners you shouldn't have, you know you're not the only one and you want to do something about it.

A useful question to ask yourself is, what would you want someone else in your field to do if they were in a similar situation? Then, perhaps you should lead by example and do it.

Another useful question is, how would you want the academic community to respond to someone else in your field voluntarily admitting to a similar level of misconduct? A slap on the wrist? An investigation? Revoking the PhD? A dialogue? Although there are no guarantees that something worse won't happen, at least you'll be able to make an argument for why you think a certain thing is appropriate.

The final question is: what are the consequences of staying silent? Are people who read your PhD papers in turn going to produce invalid results or significantly wrong opinions? Or is it niche/obsolete work that won't affect anyone much?

(Actually there is one more question: would you jeopardise anyone else's career if you admit to the misconduct?)

Instead of going public you may want to write up your story and publish it anonymously, in a place where people from your field gather, and perhaps participate in an ongoing conversation about how to improve the problem.

0

From the point of view of outside of academy: stop worrying, fix the problem, and learn techniques to work with pangs of guilt.

  • Stop worrying: these things happen. I guarantee you that 99% of all adults have some kind of skeleton in their closet that gives them a nice little adrenaline kick every few years when they remember it. Even if it's just some careless comment they said at an inappropriate moment. Shame and guilt are really common feelings that are built into all of us (or at least hopefully so). Life goes on.
  • Fix the problem: don't overthink it. Gather a few ways to tackle it, make it objective and not subjective, pick one way to go, and do it. In your particular case, although I am not familiar with the ways to get add-on articles into those publications, I could imagine you writing further papers which expound the issues from the previous one. Do not bother worrying about somehow "setting it right". Science is based on the principle of creating falsifiable theorems, and then by all means falsifying them when possible. If you really feel that you must comment on it, you can write a half sentence in your preface (e.g., "this paper fixes some systematic errors in the previous one"...). Nobody cares or should care how these errors came up or your personal story. Obviously, don't lie or make up an elaborate story about things nobody except yourself knows anyways.
  • Techniques: one technique to work on things like guilt and shame is mindfulness meditation. There are plenty of possible ways to tackle that; the usual approaches are dead simple to perform ("watching the breath"). There is nothing religious or mystical about this; the beauty is that it is fundamentally an experimental approach. You immediately see what happens, you do not have to believe anything at all. You do not need to "make belief" (this is not about imagining energy balls or stuff like that). Granted, I know people who hate stuff like this with a passion, but you might as well see if that resonates with you.

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