Found mistake in analyzing experimental data after submitting PhD thesis; what should I do?

Question

While flipping through my thesis to look up details in writing a postdoc application, I realized that I messed up my reaction rate quantification of one protein covalently bonding with another on a gel. The error I made here is that I assigned too large an intensity value for an unreacted protein control.

So here's the thing: I've already submitted my thesis to the library and none of my supervisors or examiners caught this mistake. For instance, I had to really pause and look close at the bands itself to notice that the unreacted band was actually smaller in size before going back to review my own work. Now this result was part of a bigger experiment that simply didn't work and will not be published. This was one interesting result that meant to show how two parts of a protein reacted similarly, despite one being heavier. If this result was not there, it would not upend the key conclusion of my thesis, although I would have 2/3 less of a figure.

What should I do at this point? Should I tell my supervisor immediately or file a correction? Or should I just go on with my life with the notion that no one will likely read my thesis? I feel like my advisor is glad to have me out of the lab, but I don't want to be accused of fabricating data years down the line.

Answer 1

Not to worry. Things like that happen. Tell your adviser.

Your university may have a way to add a note to your dissertation on file explaining the (small) correction. If so, take advantage of it. If not, count on the fact that it's very unlikely that years from now someone will read your thesis and draw conclusions from this data.

Answer 2

If it is such a small error, just write an extra piece of paper titled "Errata" in which you describe the error very shortly and what would be the correct thing and put it in the thesis in your library. Alternatively, put it on your homepage. Honestly, if it is such a small error, akin to a typo, no one really cares, but rather expects it to occur. NO single thesis is without these small errors! I would regard myself already as happy, if there would be not more than one of these errors per page, which makes a lot of errors for a 120 page or so thesis. Because that is, what will happen, when you read your thesis again. You will certainly find the next error. And then the next one. And another one too. And so on. So stop reading your thesis! :)

Answer 3

I am a bit worried that your response to having made an invalid assumption is to think:

should I just go on with my life with the notion that no one will likely read my thesis

What did you just spend the last 4 years (or more) of your life doing? Your thesis topic is (in theory) the culmination of years of work learning how to apply the scientific method rigorously. If your hope is that no one will read it.... What have you done with your time? Most of the professors that I know in biochemistry post doctorate work, and into the majority of the career was related to the work they did in their thesis.

Why am I worried about this?

2 Reasons.

1. What have you been wasting your life on that you hope no one will read your thesis? It should be a major accomplishment that you are proud of, that will help you as you start your career as a scientist.

2. As you apply to your first job with a PhD, Every single person you interview with will want to talk about your thesis work. It is the capstone of your PhD.

When you talk to them about your thesis, which do you think sounds better? "I made a mistake in my thesis, but I assumed no one would notice it and am glad that it is over. You should hire me because that is what I bring to the table... A person that makes mistakes and hopes no one will notice".

Or "While going over my PhD thesis, after it had been accepted by my universities library, I realized that I had made a mistake. And I then went thru the (however challenging process it is) to file a correction. I am not perfect, but I am honest, I admit when I am wrong and fix my mistakes." It becomes a talking point that you can use when someone asks you "Have you ever made a mistake and what did you do?"

What am I saying?

I don't want to be judgy... But I worry that your question was "Should I hide the mistake I made, or not" instead of "How will this mistake effect me/ what can I do to mitigate the impact it has". It makes me worried because one of these is an ethical question: Should I be dishonest and hide a mistake I made... Or should I not. And one is the question of someone who made a mistake and wants to know what they can do to mitigate the impact it has on their career.

At the end of the day, making a mistake on your thesis, while it would have been better to not make a mistake, is not a big deal. You followed your schools process and multiple people looked at your paper and didn't see the mistake. You are at the start of your career no one expects you to be perfect. But ignoring the mistake and hoping that no one will notice is a big deal... I can guarantee that your thesis will get at least cursory glance by anyone who is really interested in hiring you as a postdoc.

Answer 4

If I were you, I'd "take control" of your thesis by uploading the source files (assuming it's in a text-based source format like LaTeX) to GitHub or similar, put a label on the version that got published by the institution, then make a new commit with the corrections and put a description in a CHANGES.md file. Get the issue out in the open - this won't be the last time you make a mistake in public, and dealing with it openly & honestly, but not necessarily making a big deal about it, is a good habit to get into.

You might need to make sure you have the right to publish your own work like this, by the way, by checking with your institution.

Answer 5

There are definitely provisions like this in place at all academic institutions. This is especially the fact that your error is NOT A TYPO but a CHANGE IN VALUE which means a slightly different outcome, as you pointed out will reflect in the figure - the others on here have not recognized the type of error, but just dismissed it as trivial.

DON'T assume that no one will read it or just let it fly and hope for the best is a good attitude given the TYPE OF ERROR - as explained above

Now that you have noticed the error, write it down and what the new outcome would be had the error been captured. Then formulate this in a correction to thesis (ERRATA) section as an appendment, and then speak to your ex-supervisor to get the course of action needed to fix the submission.

Doing this is a better course of action, especially if you're planning a post-doc position that relies on credibility and honesty of academic research. period!

Answer 6

Thank you all for taking time to offer your insightful feedback. I've re-measured my bands more carefully this time and found that while a few points on my graph will change slightly but the conclusion I wrote in the text is not at all affected. So if I were to issue an errata, it would just be for the graph in the figure itself.

While some of you here have encouraged me to be proactive in reaching out to my supervisor, some of my friends with PhDs suggested I leave it for now, as the thesis is under embargo, and bring it up to my advisor when/if this figure will be published in a peer-reviewed journal. While I am genuinely concerned about how this may affect my career and want to mitigate any harm, I have heard people say that very few people read thesis and focus on published journal articles instead. What do you think of taking this approach? Does it matter when I reach out to my supervisor (i.e. asap vs. later)?

I'm not too sure what it will take for me to add an errata, but I'm not sure if my supervisor will overreact or if this may cause some drama in a department that he has already left. Please let me know your thoughts.