Well-organized research data make it possible to analyze the data not only during the project, but also several years later. That's why it's important to keep the data organized and thoroughly documented throughout a research project, from the data collection or data production to the completed analysis and processing. You don’t want to end up in a situation where you or a project colleague are unable to perform an analysis because you cannot understand the data anymore.
We have to remember that research material (data) and project planning can look quite different across projects and disciplines. That a material is in "good order" can thus have many meanings, and different projects call for different structures and different documentation. As a result, the information in this section is kept fairly general. The fundamental principle is still organization: well-organized data are data that are easy to understand and easy to find in the version you are looking for.
When you organize the project data, there are two main structures to consider: the folder structure and the internal structure of the data file, dataset, or database.
You want to have a logical folder structure, and you want folders to have as obvious names as possible. That makes it easier to save files in the right folders and to find the files you need. To make it even easier, you may want to have a readme.txt file with a description of what goes into which folder, or a description of the contents in the folder where the readme file is located. That way, everyone who uses the structure can find the right content. (See the section on folder structures for more information and examples of a user-friendly folder structure.)
In projects where there are several different versions of the data, it is important to separate the versions. You don’t want to wonder which version contains what, or what the difference is between two versions. We recommend that you document versioning rules in a data management plan, even if they seem straightforward to you at the beginning of the project. For someone else, or several years later, they may be too complicated and a possible cause for errors. A simple solution is to separate different versions in different folders in the structure.
Files need clear and simple names to make it obvious what they contain. Rules for how file names are constructed in the project should be documented in the data management plan and followed. (See the section on file names for more information.)
Good documentation makes it possible to find and understand data when they are needed. You may need this during the project or many years later. In the documentation, you explain the logic behind the file structures and naming convention in the project, but also any ad hoc decisions that may have been made. During the course of the project you should have, or be able to find, that kind of information in, for instance, a data management plan.
What is relevant to document varies between research subjects and methods, but there are often principles, conventions, and standards to follow.