academic misconduct by research student - how to move forward?

[arcturus]

I am a longtime lurker on these boards, but finally registered because I have a situation for which it would be helpful to hear different perspectives. I am tenured faculty in a science department and have several graduate students working with me on a variety of different projects. One of these graduate students gave me a draft of a paper in which the experimental set-up had incorrect information (instruments were setup with parameter x, when it was actually parameter y, etc). These values are recorded in the data logs and are in the metadata of the data products, so are very easy to find. There have been some prior issues associated with academic integrity with this student, but nothing has been documented officially.  At this point, I do not feel like I can continue working with this student, as I do not see how I can trust his/her analysis, given that the "easy stuff" was fabricated in the paper draft. I am not certain (a) how to structure the conversation about no-longer being willing to have them work on my projects; (b) what to do with/about the data they have analyzed, as I cannot trust their work, but do not have time to redo it myself, and would be reluctant to have them listed as a co-author when it does come time to publish. I have, of course, heard horror stories about student-advisor relationships that have gone south. How do I prevent such a story from developing from this situation? Is there anyway to recover the advisor-student relationship when a student has, in my eyes, committed academic misconduct related to the research project? Any advice or perspective would be appreciated.

[Hegemony]

If you are reasonably certain this is an attempt at fraud, rather than incompetence, I think you'll have to follow the official academic misconduct procedure at your institution. There is almost certainly an office somewhere that spells it all out. It probably involves you talking to the student about the problem, the student either admitting fraud or denying it, and if denying it, a hearing and then a decision. Then, if the decision is that the student is guilty, you follow the standard procedure for your department. That could mean that the student fails in part of their progress toward [whatever], or that the student is expelled. If you are reasonably certain this is an attempt at fraud, there's certainly no need to let the student off lightly, or merely to pass the student off to someone else — we could all do with less fraud in academia.

[hungry_ghost]

I'm in humanities and do not co-publish with students, but in this case that may not matter.
This is your problem

There have been some prior issues associated with academic integrity with this student, but nothing has been documented officially.

I would take the following steps -- basically, everything Hegemony said with special attention to documenting each step carefully, and making sure you communicate with appropriate administrators and go about this properly and with allies.

1. Document everything as thoroughly as possible before you have any conversations with anyone. Take time to do this right; you cannot afford to mess this up (see the part of your post I cut about "horror stories"). Also write up the prior issues as completely as possible, even if they were not documented at the time. Simply note in your write-up why you gave the student a free pass previously--for example, you talked to the student and assumed it wouldn't happen again, or you couldn't prove it, or you were so shocked you didn't want to believe it was true. If you didn't document it because you "didn't have time", well, it didn't really save you time, did it? Now you know.

2. Find out the academic integrity procedure on your campus; hopefully it will be in writing. Talk to your chair, the DGS, whoever is appropriate. Get their assessment of your findings. Take time on this; "socialize" your case.

3. If they agree that it is academic fraud, start the process to get the student removed from your lab. Use the official policy to structure the conversations. If they don't agree that this is an academic integrity issue, ask what other options there are, given that you no longer feel you can trust this student.

4. Obviously you can't trust the data they've analyzed; sunk cost. You can't publish it, so you either find a way to redo it (another student?) or toss it. Sometimes it happens, it's a waste, and too bad.

5. About how to structure the conversation with the student, that will depend a lot on #2. I typically try to have a 3rd party present. I tell the student that the person is there as a neutral party to keep the conversation fair. Actually that's only part of it. Also to keep me calm and to document everything. It may be better to have the conversation in writing, depending on personalities. But run everything by your chair first.

[Hibush]

Hegemony and Hungry Ghost both give excellent advice. Whether it is fraud or incompetence, your immediate actions are the same so you can reserve judgement while doing the initial steps. That might help quell any doubts about what you are starting in on.

Whether it is fraud or incompetence, you also do not want to continue with this student and will have to redo all the experiments before publishing. Those are sunk costs. You can grieve over that cost, but can't ressurect it. It happens in research, so it is good to accept it as normal albeit undesirable and unforeseen. Don't feel too bad about it, or look too much in the rear view mirror rather than looking ahead.

[tuxthepenguin]

Are you ready if there's an appeal? You're making a very serious accusation against a student, and what you want to do is likely to do major damage to the student's career. Every place is different, but here are some of the things I would have to deal with in this situation:

Did you talk to the grad program director and the head of your unit? This student is part of a department/college, not just your student, so you need to talk to them first. Do you have evidence in writing (email or otherwise) showing what you talked about and that they supported your decision?

Did you talk to the student and provide written feedback explaining changes that needed to be made when this occurred before?

Do you have clear evidence of blatant violations of ethical standards that would warrant what is essentially the nuclear option?

This student might be doing such poor quality work that you have to fire them. I don't think there would be any issues with that. Just make a nice, clean break from the student, and move on with your life. Once you start throwing around accusations of misconduct, the stakes are a lot higher. You better be prepared to defend your claims.

[Caracal]

This is possibly a stupid question, but I'm not a scientist. If the values are incorrect on the set up, is the implication that the student couldn't have actually gotten the results with the values they used, so that suggests they just made up the results they wanted? Or is it that the student is claiming to have done something one way, but they actually did something different?

Forgive the ignorance on the methods, but it seems like it does matter, because academic misconduct is a very serious allegation and it seems like you would want to be pretty sure this wasn't a dumb mistake. From the humanities prospective, is it more like they paraphrased something in a single paragraph without citing it, or is it more like they submitted a paper that lifted large passages from unattributed sources. The first would probably warrant a serious talk about the importance of being careful about citation, while the second would be grounds for expulsion from a grad program.

[Liquidambar]

I was surprised by the advice to report this as academic misconduct.  I just checked my school's honor system website, and we do list falsifying research as a violation.  However, everyone here seems to think of the honor system as being only for coursework.

It sounds like arcturus doesn't want to work with the student anymore.  At my institution, this would probably proceed like any other case of firing a grad student.  Once word got around, probably nobody else in the department would want to work with the student.  We'd probably figure out a way for the student to leave with a non-thesis master's degree if they were doing fine outside of research.  I'm not saying this is a great approach, but it's likely how things would play out at my school.

I'll mention, though, that my own grad student committed minor plagiarism in a course I was teaching.  This was a first year student dealing for the first time with Western culture about plagiarism and citations.  We had an honor council investigation.  The student learned from the incident and went on to be one of my better grad students.  Arcturus's student sounds much farther along so perhaps can't be rehabilitated, but if this were a newish student I might consider trying.

[Hibush]

This is possibly a stupid question, but I'm not a scientist. If the values are incorrect on the set up, is the implication that the student couldn't have actually gotten the results with the values they used, so that suggests they just made up the results they wanted? Or is it that the student is claiming to have done something one way, but they actually did something different?

Forgive the ignorance on the methods, but it seems like it does matter, because academic misconduct is a very serious allegation and it seems like you would want to be pretty sure this wasn't a dumb mistake. From the humanities prospective, is it more like they paraphrased something in a single paragraph without citing it, or is it more like they submitted a paper that lifted large passages from unattributed sources. The first would probably warrant a serious talk about the importance of being careful about citation, while the second would be grounds for expulsion from a grad program.

It is not easy to make parallels between what this student seems to be doing and misattribution of ideas and language. 
The degree of malice is unknown, but the advisor does not have confidence that the data are accurate. That is a huge problem. It could be that the student used the correct settings on the machine, but wrote down the settings the previous user had dialed in. The data are correct, but there is no way to know that. For the student to make that mistake is serious. For the student not to realize that the settings they recorded are wrong for the analysis they were doing is super serious if this is a procedure they are supposed to be expert at. Again, not malice in any of these, but reasonable for the advisor to chose to part ways.

It is also possible that some or all of the data are made up, and the student is lying about it. That is the worrisome and malicious possibility.

[hungry_ghost]

I'm not really adding anything, just pulling out a few key comments that got my attention. These are not "how to" practical responses, but responses to the bigger issue of whether to "fire" the student for incompetence or whether to tackle this as a matter of academic fraud.

First, this is the one I keep coming back to:

There have been some prior issues associated with academic integrity with this student, but nothing has been documented officially.

So, seems to be a trend with the student's work, not a one-time mistake.

If you are reasonably certain this is an attempt at fraud, there's certainly no need to let the student off lightly, or merely to pass the student off to someone else — we could all do with less fraud in academia.

Right, stopping cheaters is good, cheating hurts us all.

This student might be doing such poor quality work that you have to fire them. I don't think there would be any issues with that. Just make a nice, clean break from the student, and move on with your life. Once you start throwing around accusations of misconduct, the stakes are a lot higher. You better be prepared to defend your claims.

Right again, stopping cheaters is hard and because it's a huge accusation, you need to go about it very carefully.

I was surprised by the advice to report this as academic misconduct.  I just checked my school's honor system website, and we do list falsifying research as a violation.  However, everyone here seems to think of the honor system as being only for coursework.

Right again! Faking data is cheating.

Good luck OP.  Hope these comments are helpful.

[arcturus]

First, thank you all for providing your perspective on this. It really helps!

To clarify the situation, the data were taken correctly. The issue is that it appears that when the student could not remember aspects of the instrument setup when writing the paper, he/she chose to make it up rather than to look it up.  I view this as misconduct because they have "fabricated" the experimental setup. But, as several of you commented, it is either fraud or incompetence. Neither is great student behaviour, but one is definitely more consequential than the other in terms of the reason for leaving graduate school. In the broader context, I no longer trust that the student has completed the data analysis correctly (which is much more complicated than the data acquisition) since he/she was not able to report the "simple stuff" accurately. Given these issues, I do not believe that they should keep working on this project, or in my group.

My program has not had an advisor "fire" a graduate student in my time here, and the process is not spelled out in our department policies or in the graduate student handbook.  The graduate student is reasonably far advanced (this work would be part of his/her dissertation), so leaving my group would either mean leaving graduate school, or trying to find a new advisor (which would be difficult, as no one is likely to want to have him/her as a student, and, yes, we have discussed this as a faculty in recent times, not related to the current events).

Prior issues associated with academic integrity have been handled the old fashioned way: discussion with the student, clarification of the correct ethical behaviour, and a stated expectation that this learning experience should mean that they won't do it again.  This usually works well for our graduate students, who are usually motivated not to make further mistakes that could serverely curtail their career. This student's instincts, however, appear to lead him/her to further difficulties.  Really, in this instance, it would have taken all of 5 minutes to look up the correct description of the experimental setup. Instead, the student wrote up something that is easily provable as not describing accurately what was done.  So, the fact that the past misbehaviours were not reported to the University has more to do with trying to help educate a student without putting labels on him/her, rather than avoidance due to the extra time it takes to file misconduct reports. As I am learning, however, it does appear to be useful sometimes to have those official reports to fall back on as demonstration of past efforts to educate regarding academic integrity!

Again, thanks for your input. It has given me food for thought.

[dr_codex]

First, thank you all for providing your perspective on this. It really helps!

To clarify the situation, the data were taken correctly. The issue is that it appears that when the student could not remember aspects of the instrument setup when writing the paper, he/she chose to make it up rather than to look it up.  I view this as misconduct because they have "fabricated" the experimental setup. But, as several of you commented, it is either fraud or incompetence. Neither is great student behaviour, but one is definitely more consequential than the other in terms of the reason for leaving graduate school. In the broader context, I no longer trust that the student has completed the data analysis correctly (which is much more complicated than the data acquisition) since he/she was not able to report the "simple stuff" accurately. Given these issues, I do not believe that they should keep working on this project, or in my group.

My program has not had an advisor "fire" a graduate student in my time here, and the process is not spelled out in our department policies or in the graduate student handbook.  The graduate student is reasonably far advanced (this work would be part of his/her dissertation), so leaving my group would either mean leaving graduate school, or trying to find a new advisor (which would be difficult, as no one is likely to want to have him/her as a student, and, yes, we have discussed this as a faculty in recent times, not related to the current events).

Prior issues associated with academic integrity have been handled the old fashioned way: discussion with the student, clarification of the correct ethical behaviour, and a stated expectation that this learning experience should mean that they won't do it again.  This usually works well for our graduate students, who are usually motivated not to make further mistakes that could serverely curtail their career. This student's instincts, however, appear to lead him/her to further difficulties.  Really, in this instance, it would have taken all of 5 minutes to look up the correct description of the experimental setup. Instead, the student wrote up something that is easily provable as not describing accurately what was done.  So, the fact that the past misbehaviours were not reported to the University has more to do with trying to help educate a student without putting labels on him/her, rather than avoidance due to the extra time it takes to file misconduct reports. As I am learning, however, it does appear to be useful sometimes to have those official reports to fall back on as demonstration of past efforts to educate regarding academic integrity!

Again, thanks for your input. It has given me food for thought.

Not sure if you're still looking for input, but what you are describing is not usually what I think of when I hear "academic misconduct".

As I reflect on it, I realize that you are using the word more accurately, and that the issue is that it has become a euphemism for all kinds of other behavior: plagiarism, fraud, sexual predation, discrimination, and more. Many of these are integrity code violations (professional and institutional); many others are crimes.

Your student sounds incompetent rather than a fraud. As a result, I wouldn't write the student up for an integrity violation.

I would, however, fire the student, using exactly the argument laid out in your posts. The data collected is useless, the work needs to be redone, and there is no indication that the student can do it correctly without direct supervision, which is a waste of other resources.

[spork]

Reporting impossible results from an incorrectly-designed experimental apparatus is scientific fraud.

[Caracal]

My program has not had an advisor "fire" a graduate student in my time here, and the process is not spelled out in our department policies or in the graduate student handbook. 

Prior issues associated with academic integrity have been handled the old fashioned way: discussion with the student, clarification of the correct ethical behaviour, and a stated expectation that this learning experience should mean that they won't do it again.  This usually works well for our graduate students, who are usually motivated not to make further mistakes that could serverely curtail their career. This student's instincts, however, appear to lead him/her to further difficulties.  So, the fact that the past misbehaviours were not reported to the University has more to do with trying to help educate a student without putting labels on him/her, rather than avoidance due to the extra time it takes to file misconduct reports. As I am learning, however, it does appear to be useful sometimes to have those official reports to fall back on as demonstration of past efforts to educate regarding academic integrity!

As a grad student, my department in a very different discipline worked pretty similarly to what you describe and I saw similar problems come up, not so much academic integrity, but with students who didn't seem to be able to absorb important professional lessons. An informal approach which emphasizes learning and isn't punitive creates a much better academic environment. We all know of places where students are thrown out of programs all the time and those usually seem like unpleasant places where students absorb bad lessons about collaboration and collegiality.

The problem is that a soft approach works very well, as you say, for the majority of students who are serious, well meaning and take critical feedback seriously. In my second year I once turned in a paper for a class to a professor that was poorly edited and she buttonholed me after class and told me that sloppiness like that just wasn't acceptable in a professional setting and I needed to edit and proofread my work much more carefully. I wasn't being yelled at, and there weren't any threats or write ups. If I had been being threatened with explicit consequences, I probably would have just felt defensive and it would have been a lot less effective. As it was, I was embarrassed, but I also got the point and it was a valuable one. I also maintained a good, friendly relationship with the faculty member.

But, as you say, the problem comes when people just don't seem to be able to take onboard lessons and something more formal is called for, yet often those procedures aren't invoked.

[Volhiker78]

For your own protection, document all instances of incompetence,  save all emails and/or notes that you had regarding conversations with the student about the issues.  Especially note if you gave corrective measures that weren't followed.  All this is in preparation for any formal/legal complaint against you after you fire the student.  The same rules as for any employee termination.