Cheat Sheet on Data Organization

Research Data Management Network Data Organization Organize your data with a clear folder structure and consistent fi le names. This allows you and others to easily locate, access and use your data, to avoid duplication, and to ensure that your data can be backed up. Research Data Management Network https://researchdata.unibas.ch • This work is licensed under CC-BY 4.0. • [email protected] Illustration by F[email protected]W, modifi ed Use a fi le naming convention Use consistent, meaningful fi le names. Include the following elements: • Project/event or research team/department (abbreviation) • Author (whole name or initials) • Keyword(s) describing the content of the fi le • Date (YYYY-MM-DD) and/or version Not all characters are allowed for folder and fi le naming! Do not use spaces and/or special characters in the fi le name. Max. absolute path length: 255 characters. Use suitable fi le formats Careful selection of fi le formats is crucial for ensuring that your fi les remain accessible, interoperable, properly archivable, and can still be used after many years. Aspects of a suitable fi le format are: • No restrictive licenses • Readable by many software products (at best by open source/code software) • No encryption or digital rights management (DRM) – in case sharing is allowed • Established in the research community • Openly accessible documentation Use a folder structure • Organize your data logically in a hierarchical folder structure. • Separate active and completed work and delete any temporary fi les after use. • Don’t make the folder structure too nested, this can lead to long and complicated fi le paths. Research Data Management Network ‘How-to’ tips for data organization: 1. Do as much of the data organization as possible during the project. Cleaning up after a person has already left is usually diffi cult. 2. Have regular data cleaning days in your research group (similar to lab cleaning days), for example every 3 months or once per semester. 3. Let other group members look at your data organization to help you improve e.g. your folder structure, fi le naming, or data documentation. 4. Optimally, group leaders (PIs) enforce data clean up and set data organization/deletion deadlines. Research Data Management Network https://researchdata.unibas.ch • This work is licensed under CC-BY 4.0. • [email protected] Illustration by F[email protected]W, modifi ed Add documentation and metadata Good documentation (e.g. in a readme fi le) facilitates the reuse of data – both for others and for yourself. An important element of documentation is metadata. Metadata are structured and machinereadable information about a (data) object. Metadata are therefore data about data. Metadata include, for example, information on title, creator, date, license, fi le format and persistent identifi er. If there is no domain-specifi c or institutional standard for documentation, it is recommended to defi ne an internal project standard.