Skip to Main Content

Data Management Guide: Data Management Guidance

General Guidance

Use a standardized system for naming files and organizing folders

Document your organizational and data structure in a README.txt

Systematically back up your data on multiple media formats using the 3-2-1 rule

Convert your files to open formats when possible

File Guidance

Come up with file structures distinguishable from other projects that are easily understood by others

Describe logic behind file naming through a file naming system

  • Determine what information is important to encode in the file name
  • Establish information order for sorting
  • Create controlled vocabulary for consistency

Naming practices

  • Avoid special characters ( * . " / \ [ ] : ; | = , < ? > & $ % # ! ' { } )
  • Use underscores of CamelCase instead of spaces
  • Write dates as YYYY-MM-DD

Documentation

Enough information about the project, methods, and materials, such that the information is maintainable over time in an accessible format and valuable for those who need it

Can exist in different format types depending on disciplinary needs

  • Code commenting
  • Data dictionary / Codebook
  • README.txt file
  • Metadata schema

Provides general information about collection purpose, software needs, contact information, and how data can be reused

Describes data for future use and use by others, such as data provenance, variable names and any manipulation

Backup and Storage

Develop and maintain a consistent backup schedule

Strive for a primary copy, an additional local copy, and a third remote backup

Version files in OneDrive, GoogleDrive or manually by creating new files each time you save

Consider if the technology you're using may suffer degradation over time (CDs, DVDs, flash drives)

Convert your files to non-proprietary, open formats when possible

  • .csv instead of .xlsx or .sav
  • .txt or .pdf instead of .docx

See your LAN Manager or other local IT support for a recommendation on what network storage to use

Human Subjects

Sensitive data should be encrypted

Data with personally identifiable information should be de-identified before sharing using the Safe Harbor method

Research data should be retained in its entirety for at least 3 years after the study closes