I’m currently taking part in the Culture Track of the Mozilla Open Leaders programme, 14 weeks of mentoring and training by the Mozilla Foundation. I’m hoping to learn how to build an open culture for my lab and other projects I’m involved in.
I’ve been interested in the concept of open science/research for a while now. If you haven’t heard of open science, it’s the movement to make more of our research practices, methods, and data accessible than just the stuff that eventually makes it into a scientific journal. That includes all our dead ends and failed experiments.
I was so keen to give it a try that in 2013 I convinced my PhD student, now Dr, Hannah Read to document her experiments and data in real time in a sort of open lab book. As her project was a new area for my lab, I was hoping that the research community would offer advice as we were going along and help us avoid making any obvious mistakes. It was an interesting but ultimately unsuccessful experiment. We never got any feedback. I’m not sure anyone ever really read what Hannah was writing.
When I decided to try to give open science another try last year, I tweeted that I was looking for a mentor, and asked people to tell me how they went about making their labs/research more open. The responses were overwhelmingly the same. All variations on just do what you normally do, only in the open.
Really, that’s it?!
I found that advice so unsatisfactory. Over the past couple of years, I’ve become increasingly aware that the way my lab normally does things is less than optimal. Actually, that’s a monumental understatement, but more on that another day. Stepping into the open research waters this time around seemed a good opportunity to set us on a better path.
And that’s why I pretty much whooped with delight when I attended my first Open Leaders cohort zoom meeting. We covered the difference between things being open by default versus open by design. Open by default is defined as when someone just throws an open licence on something or just says something is open. It’s chaotic, vague, and unsupported. In contrast, when something is made open by design, it is intentional, ordered, strategic, and process-based.
Trying and failing… and learning.
Yes! That’s how I want things to be. Open by design. Here’s a good example of the difference.
In 2016 we published some of Hannah’s PhD work in PeerJ. We included the raw data from the figures so anyone could change how we plotted our graphs or potentially reuse our data. At least that was our intention… The raw data is in an excel file. With no readme file. I didn’t really even know what one of those was. Except perhaps if you saw one, you were supposed to read it? Our data makes bugger all sense to anyone except us. So much for being reusable.
Now my lab and I are learning all about data management, tidy data, and readme files, and how to manage and store our data in a way that will abide by the FAIR data principles – making it Findable, Accessible, Interoperable, and Reusable. I will definitely post our processes and what we’ve learned at some point but it’s all definitely still a work-in-progress.
In all honesty, at the moment it is much harder and more time-consuming than the way we were doing things before. But I’m confident the effort will be worth it in the long run.