Jun 28, 2023
Developing a folder structure - Working out the logic of the folder hierarchy
After the group has gained an overview of the data in the first step1, it must decide on the names for the individual folders. At this point, research groups often ask the Service Center Research Data questions such as: How many levels of hierarchy would be reasonable? Which categories should we choose? How many subfolders can we create?
Our recommendation in this case is always: “In a first step, it is more essential to develop a logic of the folder hierarchy and to communicate this logic clearly, than to follow predefined rules”. To develop such a folder hierarchy logic, we draw up a “structuring approach” for each hierarchy level and each subfolder in collaboration with the research groups.
The first level of a folder structure should exclusively contain folders with names that are familiar to every staff member (“Everybody knows this term”). Fig.1 shows a breakdown into research-relevant and group-relevant categories. Group-relevant categories are used to store administrative documents that pertain to collaboration within the group. For the area relevant to research, you should jointly decide on a superordinate initial structuring (e.g. by project or study or person).
Once the first hierarchical level has been subdivided according to group and research relevance, the next step is to turn to the newly created subfolders and to assign each subfolder its own structuring approach. To do so, you can draw on already established approaches (see Fig. 2) without being limited to them. Finding folder names based on the purpose means that the reason for collecting the data is the primary focus: For example, a work package of the project needs to be fulfilled. In this case, the folders on the first sublevel would be named after the work packages of a project. On a further sublevel, you could then structure the data according to the progress of processing, if each work package first produces raw data, then processed data, and then final results.
Fig. 3 shows an example of a practical application of the structuring approaches. Defining one structuring approach per subfolder makes it easier to navigate the folder structure, since it is clear at each level why the subfolders exist and what data is expected to be in them. Additionally, you may briefly describe the structuring approach in a small support file.
However, you shouldn't just theoretically conceive the optimal folder structure for a research group, but rather you should test it in practice. Therefore, we recommend trying out different structures and structuring approaches in meetings with your colleagues.
Would you like to update your folder structure according to this very procedure? Do you have any comments or suggestions?
We will be delighted to assist you. Please contact us at Service Center Research Data by email or book an appointment.