Pushing transparency beyond the norm
Many of the discussions that I have about transparency in research focus on the scrutiny of research methods and products. Transparency in research, people argue, is vital to demonstrate that the research was conducted in an ethical fashion and that the methods are sound. This understanding of transparency - to my mind at least - is extremely “data-centric” and focuses on the scrutiny and re-usability of data as the primary object of these actions. This data-centric understanding of transparency has undoubtedly been influenced by the rapid rise in FAIR data discussions that foreground the interoperability and re-usability of data as primary aspects of quality research.
In this project we are setting up a survey to assess data sharing practices amongst researchers on the African continent. Anelda, Thomas and I are all strong supporters of the Open Science movement, and optimising the transparency of our research and the re-usability of our data is, of course, very close to our hearts. Transparency in our research was never in question.
Nonetheless, when we came to planning how we would foreground transparency in our project design, we soon realised that we were not thinking nearly big enough. It became clear that we were missing a key opportunity to turn transparency into both a learning tool and a multifaceted tool for future research. By being transparent in all aspects of our project – not just in our data collection and analysis - we could potentially offer additional resources that could help future researchers. In particular, we asked ourselves, could we optimize transparency and re-usability in the more procedural aspects of project design to cut down the administrative burden for future researchers?
This discussion has not only been challenging, but also extremely exciting. It has pushed us to think about how we can build in re-usability elements into areas of our project that we’d never even considered. It also required us sometimes to fight against that “knee-jerk” reaction that we all sometimes encounter. The one where you have to sit yourself down and logically deconstruct why your automatic response is to not share, despite not being backed by any compelling reason. To date, these discussions have led us to commit to the following steps: Sharing the project proposal on the project website – not only to allow participants and downstream users to see what we want to do (and whether we’re getting it right), but also to be able to use the text as a guide for future applications Sharing our “work in progress” analysis of existing surveys, so that interested parties can see how we’ve got to the survey tool that we’re using Sharing not only our survey tool but also the larger question bank of questions that we developed. We we see this as a potential “plug and play” resource for future surveys Sharing our analysis code – not only so that readers can understand how we analysed the data, but also can re-use the code to analyse their own datasets gathered using our survey tool Sharing our website as a clone-able template. Many researchers lack the technical expertise/support to make a project website. We would like to offer them ours as a template if they find themselves in this situation Sharing templates for report production, so that anyone re-using our survey could potentially produce a standard report by inserting their data into the report template
We are hoping to demonstrate that by careful planning and a little creative effort we can create a range of tools that helps future researchers run surveys of their own. This is good for the researchers, and lowers the amount of set-up time they need before gathering data. It is good for us too, as (hopefully) more people will use our survey and expand our datasets. With such a win-win situation clearly visible, what is not to like?