The entity im currently at operates on the pipeline of grants=> mental health research=>publishing papers. Been with this org 1 year. Recent changes led me to have to colaborate more w/ this one coworker...

They have a background on Biostatistics, and are quite older than me but we were hired around the same time last year. I was hoping they'd be a mentor, but working with them is pain in short because I've got to put my name on papers we are submitting soon, analysis reports, etc. that I think are junk science or borderline negligence fraud. I raise concerns regarding their/our lack of understanding of statistical packages we are relying on, data integrity/cleanliness/etc.

The org is very small - only the Director could address this - but they are always convinced by this coworker during meetings whenever we express disagreements. I am not a biostats person, nor particularly gifted at all - I simply can read package manuals and see bad data/results if they're right in front of me...I am not sure if I am failing at expressing this, or if my coworker/boss simply dont understand and/or care. I find myself believing at this point that it must be the latter based on the amount of ignored concerns I've had.

  • 7
    Your question is full of terminology, acronyms and concepts that would only seem to be understood by another data analysis. This is going to be incomprehensible by the majority of people one here, and makes it hard to figure out what your actual issue is. I think from my reading that your question is simply "older colleague doesn't do what I think he should do", and "I've already had 5 jobs in the last 3 years, and don't want to quit again". You really need to edit your question to highlight what the issues are, and in a manner in that you could explain it to a non-data scientist.
    – Peter M
    Commented Dec 1, 2023 at 14:14

2 Answers 2

  • If you are not convinced about a paper, refuse to be coauthor, but agree to work in you on the datasets on the basis of being acknowledged.

  • Some people don't take you seriously in research without a Phd in the right subjects, sorry for that.


Disclaimer: I am not a bio-statistician, so most of the technical terms go over my head.

However, I can get the gist of the problem - you have very good reason to believe that someone you work with is negligently, to the detriment of the project. The solution to this is the same as it would be in any business - document, document, document. Any time you find an issue, raise it with them, raise it in meetings, and if it's not solved because your coworker brushed it off, then it goes in the excel spreadsheet with evidence of the issue and sources showing why it's a problem.

After a few months of this, you'll find one of three things

  1. The list doesn't look as long as you thought it would, so maybe things aren't as bad as you thought. It can be hard to judge these types of things since you remember the negative situations more clearly than the times when there wasn't a mistake.
  2. The list is full of evidence of "junk science", and you can present it to a superior. In this case, your attitude and presentation will be key - you are not hunting or targeting this coworker for personal reasons, you are acting out of interest for the project and company. That needs to be clear to whatever manager or director you show it to, or else you will look petty. Additionally, I wouldn't even start with your list - simply meeting with the director or whoever, expressing your concerns, giving vague details from what you documented and explain that you think there is a clear, consistent pattern of very negligent behavior. From this you can gauge interest - obviously if you are completely shut down and told not to bring it up again, drop it from there. If they show interest, you can offer to compile a list of examples, at which point you can go wait a day or two and then send over your document. From there it's out of your hands.

Keep in mind that through all of this, you shouldn't ask for permission to be great. If you have opportunities to fix problems that your coworker caused, jump on it. If your coworker tries to stop you and force their original work through, then it's just another instance you get to bring up to the director.

If you are shut down when you raise the issue, or nothing ever happens after you do, then you either have to learn to live with it or find a new job.

You must log in to answer this question.

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