Hari Sridhar

On the ethics of choosing a research question: competing labs and open science

In Part 2, Hari Sridhar continues discussions with Renee Borges on issues of research ethics in the context of the choice of research questions. 

Hari: Do you think there is value, for the science itself, in different people/labs studying exactly the same question? Maybe, as a form of validation?

Renee: This would be a waste of time if the two groups don't have completely different perspectives on the problem. Otherwise a waste of time and resources and not entirely relevant. Let's take C. elegans. Many labs study C. elegans, but you will not find two labs studying the same problem in exactly the same way. It would be inappropriate for the people involved because you would get two identical research papers, two identical streams of research being done and it wouldn't make sense. It depends on the question that is at the heart of the system, and whether you have a new completely new perspective that has not been thought of before.

Hari: What about value in terms of stronger support for a particular result? Won't two studies done exactly the same way and finding the same results give us greater confidence in the finding, i.e. greater certainty about its validity? 

Renee: It depends on whether the first study was done in an adequate manner or not. If there is reason to believe that it was "iffy" and could be improved, then yes. But again there comes the issue of ethics: if X has invested in doing a study and would like to continue in that line and then if you have another lab that wants to validate the study, refuses to believe what X has done, and wants to do exactly the same thing, that may be possible if there is good understanding between the labs. There could be all kinds of issues; e.g. X realises that it is a preliminary study and has plans to expand it, do it better, demonstrate the phenomenon better with more replicates, increase the number of sites, etc. X may be fully aware the study is preliminary and may have even declared in a publication that it requires validation and that validation is ongoing. So I don't see any point in someone else repeating the same thing. Unless you have reason to believe that something in the study was not right and it needs to be improved. Again there are ethical issues there too—the person who did the original study might wonder, why is my study not considered good enough? Why does it require re-validation? If it has problems, how did it get through peer review and get published? In general, I think it is inappropriate and detrimental to one’s own career to do exactly the same thing someone else has done. Unless it is at different points in time. For example, if something was done 50 years ago and conditions have now changed, then you might want to go back to the exact same site and do exactly the same thing and see if you are getting similar results. That is not validation but asking if findings hold true given the changed environmental conditions after 50 years.

Hari: More generally, would it be right to say that you think it's better if different scientists work on different problems instead of many scientists working on the same problem?

Renee: It depends on the complexity of the research problem. If it is a very complex problem, e.g., this whole issue of the hologenome, which is a very hot topic these days, i.e., trying to find out how the bacterial flora of your gut influences your phenotype and interacts with your genotype. Now, if one lab says the hologenome is my territory and nobody dare get into the hologenome, that is foolish. There are millions of strains of bacteria in the human gut, there are billions of humans each with a unique genetic background. This is not a problem that can be tackled by say, even 50 labs. You probably need many more labs to even begin to understand such a problem. So, it is very context-dependent and especially depends on the complexity of the problem. A familiarity with a research problem is not a good enough reason to say I will continue to work on it forever.

Hari: If I, an outsider, gets an idea for a research project after reading a paper that comes out of your lab, what, according to you, would be the ethical thing for me to do?

Renee: If it is on the same system, you should contact me and tell me you are thinking about this. If it is on a related system, it may not matter. You should ask me what I feel about it, or if I want to collaborate with you. You should check whether I feel like you are stepping on my toes, whether I already had the same idea. This is how, ideally, scientists should deal with these potential conflicts of interest. I am happy to say that it is widely practiced. I have had many such queries. In some cases I have said that I am planning to work along the same lines proposed. In others, I have had no conflict because the idea proposed was not something I was working on or was planning to work on in the future. I have also had some unpleasant situations where people have wanted to step in and the boundaries had to be indicated. It would be inappropriate for someone to start a project that overlaps a lot with someone else's ongoing project  because it is also very likely that the former's proposals and papers will go to the latter for review, with unfortunate consequences that may include rejection of the paper or proposal even though there is nothing wrong with the science. Most journals and funding agencies ask specifically about conflicts of interest when they assign reviewers. So it is best to avoid overlap, unless, like I said earlier, the problem is so complex, e.g. the human genome, that it requires many research groups to solve it.

Hari: With the move towards more open science, to greater sharing of ideas, data and protocols and more collaboration, do you think people's thinking on these issues will also change? 

Renee: In an earlier conversation we had, you had said that once a paper is published, the data and leads that come out of it are available to everyone. I don't think that these leads should necessarily be up for grabs, unless they are universal problems, which many groups could be working on to solve, e.g. the genetic code, or the structure of DNA. Even today, ethical behaviour behooves you to write to the authors of the paper if you want to work on leads within the same system. And it is so much easier now with email and the internet. Earlier one had to use snail mail or meet people at conferences, but there were also fewer groups so it was easier to keep track of what everyone was doing. There was a lot more personal contact. Today they are just names and faces on the internet.
I think the move to more open science has different motivations. One push is from people who are now focused on meta-analysis. But I also think that a large push is from journals wanting to protect themselves from fraudulent data. I am not completely convinced that it is only in the interest of the greater common good that people are starting to make their data available. Nowadays many journals are making it mandatory. There are many partners in this game who need to be protected—journals, publishers, funding agencies....not only student and supervisor. Therefore this move towards openness will not really have much bearing on what's considered ethical behaviour in issues like ownership of research ideas and projects. Each scientist or research team has a space and you have to respect that space. If I have to constantly watch over my shoulder, I may even be impelled to hurry and do bad science, or even make up stuff because I am running this race with somebody at my heels. I am not saying competition is bad but you don't necessarily have to do exactly the same thing someone else is doing. Of course, like I said earlier, in certain domains, e.g. drug discovery or working out the structure of DNA, there is a particular generic problem that has to be solved and there are many people working simultaneously on it; for some people the gestational period of the problem may be similar, and they may all be competing for that one critical piece of information to put it all together. So, one needs to decide on what is appropriate on a case-by-case basis.  But in general, it is important to be ethical, to check, not automatically assume that everything is up for grabs, to realise that people have invested in these ideas and perhaps have an advantage over you because of the greater gestation time. If you don't realise all of this, you might jump into a problem thinking you can solve it but it might turn out to be detrimental to your career, in more ways than one.

Hari Sridhar is a postdoc in the Centre for Ecological Sciences, IISc

Post new comment

  • Use [collapse] and [/collapse] to create collapsible text blocks. [collapse collapsed] or [collapsed] will start with the block closed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
  • Use to create page breaks.

More information about formatting options

This question is for testing whether you are a human visitor and to prevent automated spam submissions.
Enter the characters shown in the image.

Bookmark and Share