Meg recently wrote a nice post on what various people do in lab meetings. The small number of you who regularly read my posts will know I don’t often like bald prescriptions in statistics or about how to do science. I tend to think nuanced, context-dependent answers are better. And on the whole I think this applies to lab meetings too.
But I am going to stick my neck out and make a flat out “you should” statement. You should set aside time in lab meetings to discuss scientific ethics (and I mean discuss – students should be talking most of the time – and I mean set aside – not just mixed in among the science conversation). Inspired in no small part by Meg’s post I recently devoted a full lab meeting to such a discussion in my own lab (I hold lab meetings jointly with another professor), which I had never done before. Its not that I am uncomfortable with the topic, its just that with so many papers to read and presentations to rehearse I had never prioritized it. But Meg’s post inspired me to be creative and a little more proactive about what I thought was important. I want to say up front that I don’t believe there has ever been a major ethics lapse in either of our labs, so that is not why I did it. But in the past year I have been involved peripherally in multiple instances where various parties thought major ethical lapses were happening. And even in our own lab meeting where both advisers are very open and approachable (or so I like think at least), there was such a palpable hunger for talking about the subject that it made me very happy we had taken the time and I plan to repeat this every year or two.
Here’s how we ran the meeting. First I gave a brief verbal overview of what I perceive to be the core ethical issues (I list these below). Then I passed around index cards and made everybody write down something on the card (they could write down “I don’t have a question” if they wanted). Then I had a student trusted by the other students gather these cards. Then we opened up the floor to questions. For most questions myself and the other adviser gave our answers briefly (and we mostly agreed but differences also existed and were informative to students) and especially we let students share their opinions. And sprinkled through the conversation, the student holding the cards made sure all of the anonymous questions got posed. This obviously only works with more than 1-2 students; so if your lab is small, think about other forums like a one-time joint lab meeting, a departmental brown-bag lunch, etc. Also of note, I don’t think we walked away with definitive answers to many of the questions, but at least they were in the open and students knew what various people considered within the bounds of acceptable or out of bounds.
For what its worth, here was my brief spiel on what scientific ethics in ecology covers. I first pulled up on the screen the ESA ethical code (nobody knew it existed). Then I said there are four issues (intentionally paralleled to crimes as a mnemonic device):
- Assault and trespassing – these are ways of doing harm in the real world and relate to permits and reviews: IACUC (animal care), IRB (human subjects), environmental impact, and research permits (or private land access)
- Theft – claiming someone else’s intellectual property (ideas, data, verbage) as your own. We talked about how ideas are floating in the air and how I am much less quick to assume somebody stole my idea when I see it in a paper than I used to be. But we also talked about how there are people I trust and don’t trust with new ideas. We talked about how data ownership relates to long-term (read multiple years, multiple students) data collection efforts. I talked about how as an editor I had seen multiple cases of outright plagiarism. We talked about what earns authorship.
- Fraud – presenting as true something that is not true (specifically misleading data or methods descriptions). We talked about some sensational cases, and some less well known cases within our own field. We talked about the ethics of removing outlier data points (great if you report, fireable offense if you don’t). We talked about how you would report fraud that you suspected. We talked about how honest to be in describing methods. We talked about what to do when you are a graduate student who thinks you’ve done the analyses correctly but you’re not sure and everybody else trusts you.
- Nepotism – we talked about how conflict of interest occurs in peer review of articles and grants and we talked about what our own personal guidelines were (and how choices range from refusing to review to disclosing but reviewing to not mentioning it). Not surprisingly this was of less concern to graduate students who have not been placed in this role too often but they were certainly interested to hear how it worked.
Again the key here is this was mostly student generated. Myself and the other adviser talked less than 30% of the time. My introduction was merely a 5 minute recap of the topics above – the details I gave above of what we talked about were things that came up naturally in conversation after I stopped my monologue. And the adviser’s roles were mostly to give examples of how this was a real-world issue from personal experience and to give our own opinions. But this was primarily a graduate student driven discussion.
So even if you think your lab has no problems – no especially if you think your lab has no problems – just do it. Go ahead and schedule a discussion of scientific ethics in your lab. You’ll be glad you did. I certainly was!
What do you think? Am I over emphasizing this? Am I just slow and you already have ethics discussions in your lab? How do you do it? As a student do you feel like you get enough training in this? As a student are you aware of where else besides your adviser you go to get guidance on this topic?
I think this is a great initiative, and one all labs should adopt. Whenever I give workshops on research integrity and ethics to early-career researchers I always ask who has had any previous guidance – most often only one or two hands out of 20 or so go up, sometimes none (same applies to later-stage researchers). Young researchers in particular need the opportunity to ask questions and get advice in a non-threatening and supportive environment – about the most basic issues (e.g. why would they know that the same manuscript shouldn’t be submitted to more than one journal at a time?) to the complex and troubling. In my 20 years as an editor and as a former council member of COPE (Committee on Publication Ethics) I’ve seen many of the latter. A considerable number have been due to lack of knowledge or confusion, and might have been avoided with the right training and awareness.
Problems with research integrity often only come to light when work is submitted for publication or published. Some of the most common problems editors now bring to COPE for discussion and advice involve authorship, plagiarism, redundant/duplicate publication, conflicts of interest, correction of the literature, peer review, and data (particularly unauthorized use and inappropriate image manipulation – both of which can occur quite innocently because of lack of knowledge about what’s right/wrong, allowed/not). Anonymized summaries of the cases and the discussions are free to access on the COPE website and could be a useful resource for lab discussions.
So I agree totally with you that time should be set aside in lab meetings to discuss scientific ethics, and I’d extend this to include publication ethics.
I can’t see any responses yet – they may be awaiting moderation, but it might mean that other lab don’t have these sorts of discussions. I’d urge all to. If researchers have an awareness of integrity and ethical issues they are better equipped to make the right decisions and avoid problems.
“I can’t see any responses yet – they may be awaiting moderation, but it might mean that other lab don’t have these sorts of discussions.”
Nope, yours is the first comment we’ve gotten on this post.
Thanks Irene. I think an important point in what you say is these ethical issues do in fact arise pretty regularly. They are not some distant phenomenon that happen to other people. And while some are sheer malice and calculation a large fraction are ignorance. Time spent on education in this area is vital.
As Brian indicated in the post (and as I had in the lab meetings post Brian linked to), we have ethics lab meetings regularly in my lab. They are always interesting. It often ends up being a good way for undergrads, technicians, and beginning grad students to learn more about how science works, and it provides a great opportunity for them to bring up things that they’ve been wondering about but weren’t sure about how to ask.
So, I agree with Brian and Irene that this is something all labs should be doing. The Dudycha and Geedey book has good case studies relevant to ecologists:
and, of course, there are lots of resources on the web.
“We talked about what to do when you are a graduate student who thinks you’ve done the analyses correctly but you’re not sure and everybody else trusts you.” I do get this feeling quite often when doing analysis with new types of data, could you please detail a bit the outcome and main points of this discussion?
Hi – I think this is an extremely common feeling. It often goes under other names and frames. But I thought it was really perceptive who brought this up in the context of ethics.
I think the gist of the discussion was:
1) Science is at the frontier by definition and so its natural to have this feeling/never goes away, and indeed it is hard to do good science without pushing yourself a bit to use methods that are novel (to your or to everybody). And some occasional mistakes will be made. But such sincere mistakes are ethically completely different than intentional fraud.
2) One thing you can do is to grab opportunities to present your work before publishing (lab meetings, departmental lunch seminars, conferences). The more times your work stands up to public scrutiny the more you can be sure it is correct (or at least that the errors are small and subtle and noone else is catching).
3) When you’re specifically a graduate student, your advisors reputation is on the line too and they are usually pretty good judges of character, so if they’re trusting you, they’re probably not doing that blindly or without justification – you can take some faith in the fact they ARE trusting you.
4) None of which is meant to encourage you to rush to publication with results you are uncomfortable with. In my field of ecoinformatics and macroecology, almost every analysis I do is the first of its type. No simple ANOVA or regression for me. And one of the things I do is I “noodle”. Once I think I have a result, I put it on the shelf for a month. And then I come back to it from a different perspective and think of a different analysis to ask the same question, which hopefully produces the same answer (if not I have to work to figure out why, which are often some of my most enlightening moments). And I verify. I take a point in the end result graph and trace it back to the source data, and repeat that several times to make sure there are no errors (misaligning rows is a frighteningly common thing to do – I’ve known whole papers that had to be retracted). Before I drive all advisors trying to get their students to get their paper out the door already, “noodling” has natural limits – it is not an infinite process. Two or three iterations max is all I do.
Bottom line do your due diligence (primarily #2 and #4), and then don’t stress about it (#1 and #3).
I think I disagree with you Brian on 3) and I have different routine for 4).
For 3), why an advisor should be a pretty good judge of character (more than other people)? There is a bit of halo effect there I think.
As for 4) it is quite curious because I find leaving the results there for anything more than 3 weeks makes me deeply inefficient. How long does it take to you to get back onto the subject after having “noodled” it? Curious.
Thanks Simone. Disagreement is good!
Re #3 – I don’t think it is because advisors are gods. Its just because actually they have a lot of practice evaluating the analytical and thinking skills of other people and students. You might say its their job! Grad students are of course being trained to do this too, but unlike say math, when it comes to the variety of complexity and people, experience does count! Don’t want to go off the deepend putting faith in advisors, but I’ll stand by comment that advisors are pretty good at knowing when and who to trust to do analyses.
Re #4 – that’s kind of the point – you have to let it sit long enough for it all to fall out of your head and start it de novo (with those inherent inefficiencies but oh so important non-routinized looks). I’m not saying that about everything, because it definitely is less efficient. But for those things that are on the edge of feeling comfortable, for me its important.
Pingback: The importance of knowing and recognizing the limits of your knowledge | Dynamic Ecology