How to make your research fit for others
A contribution by Jannes Uhlott and Sophie Boße – Part 2
In the first part of the article, Sophie and Jannes explained what data quality is all about and why it is worth documenting your data as well as possible and providing it with quality information – and what this has to do with good karma. Today they explain how this works best
Data quality made easy: tidying up, documenting and using tools
In general, the topic of data quality looks very complex from the outside, but you can already prepare your data well yourself with just a few steps. The next step in approaching the topic of data quality is to document your data in detail. This enables others to assess the quality of your data for their specific purpose. You can use various tools to help you assess the quality of your data.
As a first step, it helps to sort your data. Are you actually able to sort through your folder structure yourself? Appropriate file naming with date, project name, creator and content can help a lot.
You can find more information and tips on data organization here.
You can also use version control systems to back up your data. Of course, it takes a lot of effort to tidy up existing folder systems, but don’t let this take away your motivation. Take one step or folder at a time or simply make a resolution for your new project to approach it in a more structured way from the outset. If you have any questions, you can get support from your institution’s research data management team.
When it comes to sharing your data externally or publishing it in a repository, the first step is to add metadata to the data. Try to fill this in as much as possible. This may take a few minutes, but it is essential so that others can find and use your data later – because if your data cannot be found and used, you will not benefit from the citations.
You can read an introduction to what metadata is and how it helps to document your data here or watch a video. In the CGIAR’s Data Management Support Pack you will find even more helpful information and templates for handling your research data. You can find even more information on data documentation with metadata here.
If you still have time, you can think about additional information such as the actual quality of your data. Unfortunately, repositories do not currently ask for much information about data quality. Tools such as the Bonares DQKit can help you here. Summarized statistics such as the number of missing cells or the number of duplicates as well as correlations between the variables already provide a lot of information about your data.
If you want to know more about what qualitative data is and how you can determine and increase data quality, take a look at the chapter “Quality and provenance” in GODAN’s Open Data online course.
It is also particularly helpful if you briefly explain the methodology of your work. You don’t have to write a novel to do this. For others, it is helpful to have a short (e.g. bullet point) list of the materials (e.g. sensors) you used and whether there were any external influences (such as rain or direct sunlight) that could have had an impact on your data.
Knowledge transfer made easy: how to get the next generation fit
You’re thinking: “I already do all that”? – Great! Then you can set a good example. Why not pass on your knowledge directly to the young scientists in your working group? Help them to set up a good folder structure and learn directly how quality controls and documentation can be integrated into everyday research. For example, set up a wiki for your working group in which you can share your knowledge and which can be supplemented by others at any time. This way, your knowledge will not be lost if you leave the working group at some point.
Do you want to pass on your knowledge and are looking for teaching materials on the topic of research data management in agricultural systems science that you can easily use? You can find a collection of Open Educational Resources here.
The individual steps towards good data documentation are small and don’t take much time. The best thing to do right now is to start tidying up that folder you’ve been meaning to look at for so long.
Authors: Sophie Boße and Jannes Uhlott