Directory Structure

One of the essential components of successful research data management is to establish a filing (or directory) structure for your records. Researchers are advised to structure their folders (whether paper or electronic in form) to correspond to how records are generated and that complement proposed or existing workflows. Filing structures enable research processes to be more transparent, make it easier for investigators to determine where files should be saved, and ultimately make retrieval and archiving more efficient. Moreover, established file plans demonstrate consistency and continuity in recordkeeping.

One of the most common ways to group records is by function. Because all records generated by the Harvard Medical School, the Harvard School of Dental Medicine, and the Harvard T. H. Chan School of Public Health must be retained for certain lengths of time (both to meet University recordkeeping requirements and satisfy grant-mandated retention periods), maintaining records functionally will enable retention periods to be assigned to groups of records. Like file names, folder and subfolder names should reflect the content of the folder, not the names of researchers or staff. Researchers are encouraged to document their file directory structure and describe the kinds of records that should be maintained in those folders to ensure compliance. Include basic information, such as project titles, dates, and some type of unique identifier (such as a grant number).

Some examples of functional folder titles and subfolders are below.

Administrative Records          

  • Grants files
  • Correspondence files
  • Personnel files
  • Methodologies/protocols
  • Financial records (budgets and related correspondence)

Laboratory Records

  • Regulatory records (protocols, protocol modifications, monitoring and events reports)
  • Notes and notebooks
  • Equipment records
  • Computer database files
  • Models, algorithms, and scripts
  • Statistical reports
  • Research data
  • Raw data
  • Coded data
  • Analyzed data
  • Summary data
  • Study measures
  • Audio, video, and imaging records
  • Transcripts

Clinical or Human Subjects Records

  • Regulatory records (protocols, protocol modifications, informed consent documents, monitoring and events reports, drug and device records, and institutional human research committee/IRB-related records/ethics reviews)
  • Subject recruitment records
  • Subject withdrawal records
  • Survey instruments
  • Computer database files
  • Models, algorithms, and scripts
  • Data dictionaries
  • Code books       
  • Research data
  • Raw data
  • Coded data
  • Analyzed data
  • De-identified data
  • Summary data
  • Study measures
  • Audio, video, and imaging records
  • Transcripts

Animal Research Records

  • Regulatory records (protocols, protocol modifications, animal health records , treatment records, breeding records, drug records, institutional oversight records)
  • Computer database files
  • Models, algorithms, and scripts
  • Code books
  • Research data
  • Raw data
  • Coded data
  • Analyzed data
  • De-identified data
  • Summary data
  • Study measures
  • Audio, video, and imaging records

Field Research Records

  • Regulatory records (protocols, protocol modifications, animal health records, treatment records, breeding records, drug records, institutional oversight records)
  • Computer database files
  • Models, algorithms, and scripts
  • Code books
  • Research data
  • Raw data
  • Coded data
  • Analyzed data
  • De-identified data
  • Summary data
  • Study measures
  • Audio, video, and imaging records
  • Transcripts

Within folders, files can typically be maintained chronologically, by classification or code, or even alphabetically (depending on the types of files).

You may wish to consult with the Harvard Medical School's Archives and Records Management program when establishing your file directory.

For research groups using Orchestra tools or programs, look to see if the program you require is already provided via modules (https://wiki.med.harvard.edu/Orchestra/NewUserGuide#Available_Software_Environment_M). If you do not see the program(s) in the modules list and the program(s) that will be run by multiple users, then the best practice is to establish a shared directory, such as “/bin”, within the lab directory for storing these programs. Programs stored in a shared directory can be accessed and run simultaneously by multiple users.

 

Last Updated: 2016-12-20