9

Currently I'm affiliated with two universities: Nankai University in China (postdoc) and Monash University in Melbourne (adjunct). I'm attempting to apply for ethics approval for a research project involving personal data that people post online. I'm completely new to this.

Q: Do I need ethics approval from both universities?

It would be substantially easier for me to get approval from Monash, simply because the forms are in English and the people I'd need to discuss it with speak English.


What actually happened: I applied for ethics approval from Monash. They were fairly patient with me, being new to the process, and it took a bit of back and forth. Nankai University matched the conditions and dates by Monash, so I didn't have to bother with filling in paperwork in Chinese. (And the paper was subsequently published here.)

  • 1
    Where do those online posters come from? And can those data potentially reveal the poster's identity? Aka, will you get names, IP, address, etc.? And finally, which one are you primarily affiliated to? – Penguin_Knight Apr 9 '15 at 0:22
  • Are you going to list both affliations on the publications? Are you afraid of getting sued? – StrongBad Apr 9 '15 at 0:40
  • The data are user comments to news articles. Sometimes it uses their real name, and links to their Facebook page (which can give you a lot of personal information, even without being friends); no IP addresses. It may be that some of these comments are by minors. I'm primarily with Nankai University. I'm not afraid of being sued (there's nothing sue-worthy), I'm afraid of my paper not being reviewed by a journal if I don't get approval. – Rebecca J. Stones Apr 9 '15 at 0:42
  • 3
    Almost certainly you have to at least ask each one whether you need their approval. If it turns out their policies say that you do, and you went ahead without it, telling them "someone on the Internet said I didn't need your approval" will not save your job. – Nate Eldredge Apr 9 '15 at 2:11
  • 2
    @RebeccaJ.Stones We do not and cannot know the answer to your question because it depends completely on your universities' policies. You must figure out how to ask them. – David Richerby Apr 9 '15 at 8:20
6

The purpose of IRB approval is for four things. The first is to make sure research is being done ethically and responsibly. Except when used in a circular definition, research does not need IRB approval to be conducted in an ethical manner. It definitely does not need multiple IRBs to approve it. The second is funders may not fund research that is not approved. They would only want approval from the institution where the funding is being given to. To publish research, it general needs IRB approval, but again, they do not care where it comes from. From both the publisher and funder point of view, IRB approval is their proof that the research was conducted in an ethical manner.

The final reason is so you are not liable. IRB approval, at least at the universities I am familiar with, means that the university takes on the liability as long as the protocol is approved. As place that may get sued,more any place that you will want help from if you get sued, will likely need to approve the research. Many IRBs have a light touch review for studies that have been reviewed by another university's IRB.

  • This is good: it answers the general question. In my case then, I think it would be reasonable to get IRB approval from Monash alone. There's no real chance of suing, and the research is effectively unfunded. – Rebecca J. Stones Apr 9 '15 at 1:18
  • 1
    @RebeccaJ.Stones Whether or not you get sued has nothing at all to do with whether or not you are funded. Also, why only the concern with getting sued? You could lose your job if you don't get proper clearance. – David Richerby Apr 9 '15 at 8:27
  • 1
    You are NEVER not liable if a subject gets hurt. The ethics review process and informed consent has little to do with liability. A good informed consent design can make it very difficult to claim that a subject was not made aware of their risks, or at least anticipated risks. If a subject gets hurt from an unanticipated risks, though, or from a violation of the protocol, all bets are off. I suggest removing the bit about liability. – Scott Seidman Apr 9 '15 at 12:41
  • 1
    Generally, the answer will depend on the laws and practices of both your employing countries and employing institutions. It is not simply a matter of funding, as I have had to do IRBs for multiple universities for unfunded pilot projects. (US) It would not be wise to neglect checking with your Chinese university's ethics board. I suggest editing this answer to remove the reference to not needing multiple IRBs. Also, the liability portion is potentially misleading. Some new device research for some universities can require assumption of liability by the researcher, even if passed by IRB. – erwin Apr 9 '15 at 14:42
10

Yes.

Or at the very least, check with both Universities' ethics boards.

I know of one instance of a post graduate student who in a similar situation got clearance from one university, but not the other, and was forced to discard their results.

  • +1 - The only people who can answer this definitively is those ethics boards. – Fomite Oct 31 '17 at 3:15
2

In the US, you generally need the approval of an Institutional Review Board (IRB) if your work was funded by the federal government, and you would have to demonstrate that approval to publish in any venue that required it. There are probably many other institutions and laws in the US that require IRB approval for this kind of research, but most of the issue is driven by the federal government's attachment of the process to its research funding. If this were happening here, either one should be sufficient. That being said, if the work is mostly related to your postdoc, I would be most concerned about the requirements of the funding agency or university that funds it. An adjunct position in the US would imply mostly a teaching role to me which would be less concerned about your research work. Your situation may be different.

2

You first need to know the laws of the country where you "are". In the US, as long as you're not engaged in directly-regulated research (such as medicine), IRB approval is imposed on you via a contractual / employment relationship with an institution. Therefore, in the US, you are required to do what your employer requires of you. An institution may require that all research conducted using institution resources undergo IRB scrutiny. In the US, the institution would have no power to limit your independent research activities, but such assumptions of individual liberty may not hold elsewhere. Other countries may have other laws; if you're actually in China, you have to do what Chinese law requires of you, even if you're only occasionally in China. It does not matter, from the legal-enforcement perspective, whether it is difficult for you to pursue this question in Chinese, following prevailing cultural norms. You should therefore find someone who can give you honest and expert advice about IRB law (civil or criminal) in China. And just as it is in the US, you need to inquire of your employer what they require you to do (where the consequences of violating their rules could be getting sacked).

If you are only concerned with publication issues, you should inquire directly of relevant journals what their specific requirements are. Some journals do not raise the question at all (I only know by rumor that it's required in psychology). If a journal requires you to warranty something about IRB approval, you need to know in advance exactly what is required. While I assume that approval from Monash would suffice, you really should verify that that is the prevailing policy for journals in your field. In the worst case, if some journal requires IRB approval from each institution where you are employed while conducting the research, they just don't submit your work to that journal.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.