Include open science statements in ethical approval applications: an open letter to researchers and ethical committees

Include open science statements in ethics applications:

An open letter to researchers and ethical committees

tl/dr; I ask that ethical committees require that researchers make a statement about the sharing of data generated in a research project. I also ask that researchers do this anyway. Oh, and I include a few to many Harry Potter references.

like this one

Open science is awesome. Near the end of 2017 I shouldn't need to bang on about how necessary open science is or how shitty the alternative of locking away research outputs so that only the special people with the right access can look at the interesting things (I'm looking at you academic publishers, your ridiculous journal bundles, and your business structure that any online scammer would be jealous of).

What I love about the open science movement is the fact that researchers themselves want to break down the ivory tower of research institutions and share their work with everyone. It builds transparency and raises the bar for science. It gives anybody with access to the internet the capacity to learn about current research, why it is important, and how it is done.

New book idea: Harry Potter and the Open Science High Five

For the most part, taxpayers and philanthropy fund science. They pay for the equipment we use, the data we collect, and they pay researchers wages to play with that equipment and data. The fact that research outputs in the form of journal articles are often behind publishers paywalls to the very people that funded that research (and the publication costs) in the first place is a cluster-fuck of awfulness.

All of the blame cannot go to the publishers or journals, the fact that the materials and data generated from research is usually hidden is another problem. Why? I hear somebody asking from the void.

Here's an analogy borrowed from the Black Goat Podcast (FYI they're awesome, give them a listen). You trust your boyfriend/girlfriend/significant other, they are nice to you and they have interesting things to say, but, they spend a lot of time on their phone. Say that every time that you are in a position to see the screen, they lock the phone and move away from you. Eventually you might get suspicious and ask kindly what is on their phone that they do not want you to see. If they show you without a hassle, then your fears are relieved and maybe they were actually using the private browsing tab to find you a birthday present. Great, problem solved. Alternatively, if they tell you to mind your own business and get fucked, your trust in them is hardly likely to improve. No, you're going to suspect that something terrible is going on behind closed doors.

Hiding data and black boxes of analyses procedures does not inspire confidence in your research.

You know what does inspire trust, if your partner leaves their phone open on the table for anybody to see. Open to check, because they have nothing to hide. Wait, we can do that with our raw data and analysis scripts, say waaaaaaaaaaaaaaat. Yes, open science is awesome, and data/materials can be shared easily online.

I also had a Harry Potter and the Chamber of Secrets analogy. If you're data are hidden and protected by a huge ass snake behind a door of snakes, you are also not inspiring confidence in your hidden practices.

In the analogy, Ginny was the data left to die in a dungeon. Too graphic?

An easy proposal to ethical committees and researchers

Here is my small proposal.  It will take a little work on behalf of ethical committees, but I believe that it would have a wide impact on open research practices. At the very least, it will force researchers to think about open research. It will not take more than a few minutes effort on the part of researchers applying for ethical approval for their next research project.

Ethical committees

You give us forms to fill out before we do research. We moan about them and doing more paperwork, etc. Eventually you assess our research plan for ethical implications. Sometimes our moans are valid, like removing the word "experiment(er)" because it conveys an impression of "human guinea pigs". Most of the time, however, the issues raised are valid and important, including ensuring adequate information and debriefing is given to our human guinea pigs participants.

the three stages of writing ethics applications

Catching researchers on ethical oversights is important. So, you could also catch us on how the research outputs will be distributed. One massive ethical oversight is publishing research in a journal that requires data sharing as a condition of publication and then refusing to share the data (special thanks to James Heathers on Everything Hertz Podcast for raising this point - also highly recommended listening).

There is usually already a section on anonymised data, that could easily be added to. All it needs is another box in the application "Please provide a summary of how the anonymised data will be distributed, e.g. to an online data hosting service. If the data will not be openly shared, please state why and include a description of what conditions need to be met for another researcher to access the data.

I already have to include a statement about publishing the results as a paper or in my Thesis as part of the participant informed consent forms. So, what about a statement like "I understand that data that I provide for this study may be shared openly on a platform such as the Open Science Framework. I also understand that the data will be anonymised and every attempt will be made to ensure that my data cannot be linked to me in the future".

The potential ethical issue is that, without a statement like this, is it OK to share the data in the first place? For me it seems to be a grey area. One problem is that feasibly, there is no way that researchers could recontact all participants to try to get consent to share their anonymised data

Researchers

If you have published a paper under the condition of sharing of data, then the data should be shared. You already made that decision when submitting to that particular journal.

Just do what I suggested above, please. Give a summary of how the data, materials, and analysis scripts from the research will be shared in the ethical application. Think about these issues before you even start the study and leave yourself open to share data in the future. Let participants know what you plan to do with the data. They might even like you more if you say that you will make the research available to the public.

There are valid reasons for not sharing data. For example, the population studied was so specific and the data so detailed that anonymity cannot be guaranteed. In this case, discussing this before the research has been conducted and having a record of why you will not share openly (but will do under a direct request) can only be a benefit to the transparency and honesty of your specific case. At least then you're not freaking out your partner by always hiding the phone, "I'm searching for a birthday present, you'll unwrap it soon!".

Journals

If you have a statement that requires data sharing, maybe ask for a link to the accessible data when the paper is accepted. Or, require that the data is submitted or archived in a specified way alongside the initial paper submission. It's hardly a condition of publication if a moody researcher can still tell people that they would sooner throw bricks at them than share their data.

The take home

A requirement to share data probably cant be added as a condition for institutional research ethics approval. There would be too many complications, winging and genuine problems with that approach. What can be done is including a statement about where, when, and how the research outputs (papers, data, code, etc) will be shared, and if not, then why. We need more of this kind of transparency, and it is exactly the kind of ethical issue that should be overseen by ethical committees.

From now on I'll have this kind of ridiculous image in my head when hearing stories of people flat out refusing to share data when they have completely agreed to doing just that.

Also, sorry for getting carried away with Harry Potter references