Research Data Services at Penn State Harrisburg

Practicing good file naming behaviors can save a researcher time and headaches when searching through data files. File naming is a framework for naming your files in a way that describes what they contain. Having a clear naming convention – before you begin – helps you stay organized (and it’s easier!). Remember that your file naming convention should be consistent between research collaborators. Below is a list of file naming best practices to follow:

• Meaningful but brief names,
• Most important information first,
• Avoid using spaces, dots, and special characters,
• Use hyphens (-) or underscores (_) to separate elements in a file name,
• Use a specific date convention when possible (YYYYMMDD OR YYYY-MM-DD), and
• Use leading zeros for sequential numbering.

Version control clearly distinguishes between individual versions of specific files and is important to research projects both big and small. Proper version control practices ensures the ability to track any differences, prevent unauthorized file modification, and prevents the loss of information. At its best, it preserves data authenticity. Below is a list of version control best practices to follow:

• Keep an original “master” file for raw data,
• Determine how many versions you will keep and for how long,
• Maintain a file history, version control table, and/or additional notes included with your files,
• Document changes between versions in a separate file,
• Include date and/or version number in the file name, and 
• Distinguish between versions shared by researchers and working versions of individuals.

Structuring your data folders in a clear and consistent way makes it easier to locate your data. Good practices make it easier to organize files and version. All relevant material to the data should be included in the project folder. Included in the folder should be detailed information on data collection and/or processing procedures, if that is relevant to the project. Project folders can be structured by activities, by data types, and/or by materials.