Research Data Services

File Naming

Establishing file naming conventions and folder hierarchies at the beginning of a research project will make it easier for you and your collaborators to navigate and find specific files, and avoid file duplication and accidental deletion. The most important rule is to be consistent. Best practices include:

• Keeping files names short but meaningful
• Using capital letters to delimit words, avoiding spaces, hyphens and underscores
• Avoiding the use of non-alphanumeric characters
• Denoting dates using the "YYYYMMDD" format

For further guidance and examples of best practices, see Stanford Libraries guide.

File Organization

A well structured file hierarchy will make it easier to locate and share your files. Some recommended practices include:

• Restrict level of folders to three or four deep
• Consider limiting the number of folders within each folder to ten
• Include a folder within the folder structure for “documentation”. This might include:
  • Project proposals/protocols
  • Consent and approval forms
  • Methodology documents
  • Data management plans
  • Code used for analysis and outputs
  • Codebooks or guides

See an example of folder hierarchy from the UK Data Service.

Documentation

For research data to be read and interpreted correctly, it requires sufficient description and documentation. Consider the information needed to make your data ‘independently understandable’, now and in the future. For example:

• Documentation of experimental methods
• Software code used for analysis (see related guidance)
• Structured metadata
• Readme Files (see Cornell University template)
• Codebooks / Data Dictionaries (see OSF guidance)
• Variable descriptions

Remember that it is best to document your data throughout the research process, rather than at a later stage.

File Formats

Typically, file format selection will be determined by the software you use for data collection or analysis. However, because technology changes, storing and sharing research data over the long-term requires it be kept in a format that is widely accessible and readable. Researchers should also consider whether their data needs to be converted to an open file format for archiving once their project is completed. 

Some examples of file formats:

• Statistical data files e.g. SPSS, SAS, Stata, R
  
• Spreadsheets e.g. Excel or Google Sheets
• Text data e.g. .txt text files, .csv comma separated values, etc
• Documents e.g. MS Word, Google Docs, PDFs
• Image files e.g. .tiff, .jpeg files
• Geospatial files e.g. Esri Shapefiles, GEOJSON, etc.
• Scientific files e.g. .NetCDF, .MAT

For recommendations on file formats consult these resources:

• UK Data Archive. Recommended formats for data sharing, reuse and preservation.
• Library of Congress. Sustainability of digital formats.
• Library and Archives Canada. Advice on the digital file formats to be used when transferring information.

Metadata 

Metadata is "structured information associated with an object for purposes of discovery, description, use, management, and preservation. Metadata are often called data about data or information about information." (NISO, 2004)

Metadata should follow defined standards in make it findable, accessible and reusable. These standards vary according to the research discipline. For help identifying an appropriate metadata standard for your project, contact UVic Libraries. You can also search for metadata standards in your discipline through the DCC open directory:

• Disciplinary Metadata -- DCC

  The DCC maintains a list of disciplinary metadata standards

• RDA Metadata Directory

  The Research Data Alliance's Metadata Working Group is developing an open directory of metadata standards applicable to scientific datasets