Problematic ethics

scout.jpg

Human studies

Ethical considerations

An accepted guideline to start with when doing human studies is the Helsinki Declaration. When you do human studies, you must:

  1. Make an application to a regulatory body describing your study in great detail. In Singapore, the regulatory bodies are the Institutional Review Boards. You can start a study after your application has been approved.

  2. You application has to describe exactly what you are going to do in the order you are going to do it in, why this study needs to be done, what you hope to achieve and when, and other very specific details.

  3. You must take permission from the the individuals taking part in the study, the participants. You must explain everything they care to know about the study, before and/or after they've taken part. You really should make this explicit in the publication produced from the study, as in 'Participants took part in the study after it was explained to them and they signed an informed consent form' (sometimes in psychological studies we need to lie to the participants before the experiment, then we tell them the truth after the experiment - obviously that is not relevant if it is a study on cancer for example). The participants should have the right to receive any publication that will ever be produced from any data to which they contributed.

  4. When you receive approval from an Institutional Review Board to proceed with a study, you'll get an IRB number. You must write this number in any publication produced from your study.

  5. The rule of thumb is one application, one study. The means that you get one IRB number to do your study, and you put that one number in the publication produced from the study. If the study is big and more than one publication is to be produced from it, you'll need to think about how you'll handle the statistics (see Statistics below) but there is no discussion about the fact that you have to put the IRB number in all the publications produced from that application you made. Why wouldn't you, if you're being ethical and honest?

The ethical problem

Researchers in the Dementia Consortium and Singapore may write in a publication that they've complied with the Helsinki Declaration and Institutional Review Board regulations. But then they do not put in the number. Or the professors do not say that they took permission from the participants, or tissue was obtained from one place and analyzed somewhere else, so we really need to know what kind of informed consent the participants signed but we do not have that information. A professor at NTU wrote the same number in several publications in similar studies as if they were different, which they can't be if the number is the same. Actually, when confronted, that same professor at NTU confessed he used the same data in different publications and did not tell us that he did. It was not really a confession because he said it like: 'So what? What's the problem?'

Animal studies

Ethical considerations

  1. Before you start a study using animals, you need to apply for and receive approval from the Institutional Animal Care and Use Committee. In theory, that application is also reviewed by some kind of government regulatory body, in Singapore that is Animal & Veterinary Services (see Who is in on it in Frequently unasked questions). You need to explain what you are going to do in the study, how, when, why it is necessary to use animals and what the study will contribute, why you will do the study the way you say you will, how many animals you intend to use, and importantly the steps you will be taking to minimize pain and suffering in the animals. Basically, the purpose is to secure the wellbeing and safety of the animals and researchers, and the successful outcome of the project.

  2. The laws, policies, and guidelines which control animal experiments are by and large similar from one country to another (there are important differences). In Singapore it is the NACLAR and GMAC Guidelines regulating animal experiments, under the Animals and Birds Act. In those Guidelines, similar to others, there are 'alarm bells' which go off and which mean we need to call in all kinds of scientists, veterinarians, laypersons, and experts to discuss the matter. The loudest alarm bell is an epidemic in the animal facility. The second loudest alarm bell is if you say you will be killing animals without anesthesia. You really need to have an indisputable reason why you need to do that. Let me tell you: it is extremely difficult to have any defensible reason, good or bad, why you should want to kill animals without anesthesia.

  3. When your application to do animal experiments is approved, you receive a number which you use for paperwork and which you really should write in a publication produced from the study. Again, one application, one study. In Singapore, this number is called Animal Use Protocol.

The ethical problem

Vyas and Mitra at Nanyang Technological University kill animals without anesthesia for no scientific reason. It appears they have been doing this for quite some years, and after they were reported. They do experiments that cause significant pain and suffering for the animals. I do not know why. Possibly to maintain a pretence of research, but a pretence of research can be maintained without effectively torturing the animals, and traumatizing the researcher in the process.

Other researchers in the Dementia Consortium very rarely write the Animal Use Protocol in their publications. Often when you do see a number, the number is not for studies done in Singapore, it is the number from an Institutional Animal Care and Use Committee abroad, such as the States or China, which tell you that other countries are more concerned about this. Sometimes it is not clear in which Singapore institute, or indeed in which institute in the world the experiments were done, because the Dementia Consortium professors say their experiments were approved in this institute, that institute, this country, that country, and those recommendations there. This is nonsense because one set of experiments can only comply with one set of guidelines. And then we do not know where the experiments were done. We know that animals and researchers are different from one facility and institute to another, no matter how hard we try to 'control' for all controllable factors. So if you want to do a set of experiments that actually produce a meaningful result, make some kind of sense, you really should do the whole set of experiments using animals from one facility, in one institute, carried out by one group of workers or researchers.

Affiliation

 

Ethical considerations

  1. Let's say you're with University X and you publish an article. Obviously, your affiliation on that article will be University X, where you did the work and might still be working. Say you moved to University Y after you finished some work at University X. What should your affiliation on the article produced from this work be? Clearly, University X has a right to be on the article. Maybe you finished the work later, so University Y has a right to be on the article too, because that's where you got paid while you finished the work. University Y is in any case your present address or affiliation so it has to be on the article so we know where to contact you. Usually there's no problem, and there is plenty of room for you to be honest about where you did the work, who contributed what, how much money went into the research from University X, Y, and anyone else. Also obviously, we should be able to reconstruct your career trajectory from your articles: first you were at University X, then between X and Y, then Y, so on.

  2. Once upon a time, a professor had an affiliation. Then it became fashionable for hot-shot professors to have two, three, or more affiliations. Obviously, you can't be in Sweden, Brazil, and Singapore all at once, or at NTU, NUS, and NNI (three institutes in Singapore) at the same time and supervise projects in all those places, mentor students, and do everything a professor needs to do. If you walk into a doctor's clinic or a lawyer's office and see a wall covered in certificates, you should leave. The same with professors affiliated with numerous institutes, consortia, boards, so on. Do not be intimidated, it is just marketing and politics, not honor and expertise.

The ethical problem

When it comes to professors in the Dementia Consortium, throw ethical considerations 1 and 2 out. Trash them, they do not apply. The same professor might list a dozen affiliations in just two years. These affiliations might be at different institutes (University X vs. University Y) or at the same institute (Department Z at University X vs. Department W at University X). I am not sure why they do that. Might be for the paper trail following the money - this money traces back to that institute, that money to somewhere else. Might be to publish the same work in more than one article (see Duplication) as if it were done somewhere different. Could be just smoke and mirrors out of habit?

Funding

Ethical considerations

  1. Let's say you're a doctoral student in a medium sized group at University X. You get your salary from Graduate Program Y. Foundation Z is paying for the cost of the project you and your group members are working on. University X is paying for a for fancy microscope at a Core Facility which you are using. When you publish an article or your doctoral thesis then in Acknowledgements you need to write: I got money from Graduate Program Y, Foundation Z paid the project costs, and University X paid for the Core Facility and I'm grateful to the staff there.

  2. In theory, one applies for funding for a project, submits updates as the project is going on, and a final report after it is finished. A very nice grant for a medium sized group working in life sciences might be enough for around five years. In practice, we don't know how things will go because we are in the business of exploration. We can guess what the cost will be, but really we don't know how much it's going to cost; it depends on how things go. And we need the money now, not later. So what often happens is a well-run medium sized group will have a clever coordinator who manages a 'pool' of money used for several projects. At the end of the day, the ethical thing to do is to transparently acknowledge who paid for the project, and be honest with the funding bodies about what happened with the project you said you would do.

  3. Grants have numbers too. You must mention the funding body and grant number in the acknowledgements.

The ethical problem

Horrific. Members of the Dementia Consortium are very often not only extremely generously funded by the Singapore government, their publications frequently mention generous funding from abroad. But the grant numbers do not match, they do not add up. For example, we may know that grant number so-and-so is for a project on liver cancer, but the professor writes the grant number in a study on epilepsy. In the bigger picture, looking at the generous funding received by Dementia Consortium members and the work actually produced tells us that, simply, money is disappearing. There is nothing to show for it, or what can be shown for it does not match the grant, or is published much, much later than when it should have been.

Statistics

Ethical considerations

It's complicated but there are simple rules which every researcher should know, such as you must not start statistically analyzing your data before you're done collecting it (or else you might try to 'influence' the data while you're collecting it). The way I think about it is how a Professor of statistics explained it to me, that statistics is a kind of magic. It can be very powerful or it can be just tricks, illusion, cheating, and charlatanism. You can use the simple spells, but if you need to use a complicated spell, then go to the Temple of Statistics and talk to an initiated priest before you start the work, while you're doing it, and after. Or else risk the spell blowing up in your face and you spend the rest of your life as a frog. At the end of the day it comes down to the researcher: do you want to answer a research question in a valid manner, or do you just want to prove your lovely hypothesis is right and publish a paper, and bend or break the rules to do so.

The ethical problem

I am not initiated into the Temple of Statistics but I know foul play when I see it. And it is in plain sight in the work of some members of the Dementia Consortium. For example, see Vyas and Mitra.

Duplication

Ethical considerations

Let's say you are working with a group on a project. You want to get your PhD and so does someone else in the group. You want an article with your name first in the author list, and so does your co-worker. At the same time, your professor needs to have at least three articles published this year or the professor does not get tenure or has much less opportunity to get that big grant - anyways the professor wants lots of articles. What can you do in the notorious publish-or-perish environment? You do salami publications. You have one roll of salami, the project. You slice it up. You publish a little bit of the project using Dataset A in Article 1 with you as first author, another slice of salami using Dataset B in Article 2, and a third slice using Dataset C in Article 3. Everyone's happy and if no one catches you, you're good. Even if someone does catch on, it's hard to pin it on you, and hey, what can we do if we have to publish, publish, publish, or be 'eliminated'? Duplication or salami publications are not OK. Notice you also run into other problems with salami publications, for example, with ethics approval and statistics.

The ethical problem

Outrageous. Duplication appears to be standard practice for many or most members of the Dementia Consortium. And it gets worse. You may not use the same dataset, the same slice of salami, in more than one publication, of course. If you do, you have to say you did and why. Members of the Dementia Consortium not only practice salami publications, they appear to duplicate and combine salami slices. In other words, a professor in the Dementia Consortium may use Dataset A in Article 1, Dataset A again and Dataset B in Article 2, Dataset A yet again and Dataset C in Article 3, and so on.

Regulation by Singapore institutes

For all the problems mentioned above, there are regulatory procedures in place at Nanyang Technological University and National University of Singapore to stop them from happening. For example, animal experiments should be very carefully reviewed and discussed before approval is granted to start those experiments. Furthermore, Animal & Veterinary Services in Singapore should be keeping an eye on that process (see Frequently unasked questions). Obviously, this system failed.

Anyone working in these institutes should upload their work onto the University repository. The University has theoretically very strict instructions on how to report funding, affiliation, and ethics approval numbers. Well, it appears one can ignore all that. Or there are systemic ways to obfuscate (muddy the waters). For example, researchers at Nanyang Technological University mix up the author list when they are uploading their publications on the NTU repository (DR-NTU), or upload one article onto the repository and not another (duplicated) one.

What is very worrying indeed is that regulation of this regulation also failed. Think that in other institutes, say in Portugal or Germany, professors have been fired for 'self-plagiarizing', copying a paragraph or two from one of their publications into another. If you think that is something like petty theft, then what is going on here is murder. I reported to Nanyang Technological University and they just dismissed everything in a one-liner. I sent a report and reminder to National University of Singapore - they should have acknowledged that they received the report, whether or not there will be an investigation. They did not acknowledge.

This is important because if you are working or studying in Singapore, and are told to engage in illegal or unethical activity, you will not be able to report it. See What this means for you.