UO Libraries Home Page

2.1 What's in a recordkeeping/data management plan?

Below are some tips that provide further information and guidance on data management and records preservation during the research process.

Data management plan

  • A data management plan is a map that guides a research team through the process of organizing, documenting and controlling the information that is created and managed during and after a research project. It spells out what to manage and how to manage it.

Components of a plan

  • File structure (a scheme that organizes files in a systematic way)
    • Develop a file structure for the data as well as the records created during the project
    • Make sure to document in the file structure the location of any files that reside on other servers, computers or locations. Relationships between all records and data from the research project should be documented.
  • Naming conventions (rules applied to naming files in a project so that they are easily identified)
  • Data integrity/quality assurance
    • Assign responsibilities for each master file to specific research team members to lessen potential inaccuracies
    • Establish version control of files through a naming convention and document changes to master files
    • Maintain old master files during the research project in case newer versions become corrupt or contain inaccuracies
  • Preparing record and data documentation
    • It is ideal to have metadata (data about data) down to the variable level
    • See the Data Documentation Initiative for information on technical metadata requirements
    • Link analogue and digital research materials in documentation
  • Project documentation
    • Document decisions and steps taken in the research process
    • A simple way to document the project or specific data groupings is to create a "read me" file at a root level (e.g. in the main folder of a group of files) of a grouping of files and describe what the files are, how they are organized and what has happened to them
  • Backups and data storage
    • Determine frequency of backups for the project
    • Validate backups with checksum algorithms
    • Store backups offsite
  • Data security
    • Ensure network and computer security - protect with passwords and do not store confidential data on servers that also host internet services
    • Physical security - lock the room where confidential data is stored and restrict access to the space
    • Keep hardware up-to-date with patches and virus protection; otherwise machine can become vulnerable and potentially compromised
  • Data disposition
    • Work with a data librarian or archivist to determine if the data should be permanently preserved or if the data should be maintained for the time required by federal, state or institutional policies and then confidentially destroyed.

Elements of good datasets

  • Establishing each of the elements above should enable you to create reliable datasets that will be appropriately documented and interpretable to another researcher in the future.
  • Other tips for creating a good dataset
    • Double or "blind" data entry can help eliminate inaccuracies by catching discrepancies by re-entering the data.
    • Sample a small percentage of the data and check it for accuracy.
    • Limit some of the complex data manipulations or organization to one or two people on the project to limit inaccuracy or inconsistencies.
    • Check for any portion of the dataset that seems extremely variant - small inaccuracies like that can cause major shifts in data analysis results.
Go to 2.2 Develop Data and Records Workflows
Maintained by: Erin O'Meara, erino@uoregon.edu
Last Modified: 11/11/2007