What and How to Preserve

Preserving research data goes beyond storing files in a secure location. Following best practices for the preservation of your data files allows for future reuse, by yourself or others. It also demonstrates a commitment to maintaining the scholarly record of your research, and can be an important consideration by funders of future projects.

Selecting what to preserve

For most projects, you should be selective about which files are kept for the long term. If you have file storage costs or limits, you may already have a plan for which files you retain to support your final results. Some things to consider to help you select which files you will preserve:

  • Do the data directly support any published articles or reports?
  • Can the data be reproduced, or were they generated under unique circumstances (e.g. observational data collected in the field), or at great cost (e.g. purchasing computing time on high performance machines)?

File formats

All digital files are vulnerable to becoming obsolete with new technology, but if you save your data in accessible and open formats, you reduce some of the risk of losing access to your data over time. Helpful guidelines about file types can be found at:

Documentation/metadata

Thoroughly document your steps—while collecting or generating your dataset, the relevant processing steps and analysis methods you use, and any other information another user would need to understand your data.

For more information about data documentation and metadata, see the page on Describing Data.

Selecting a repository

Depositing research data with a trustworthy repository is a key step in preserving your data for future use. There are many data repositories to choose from; look for one that is suitable for the type of data you’ve collected, and follows the principles of sustained access to research data.

For more information about selecting a repository, see the page on Data Repositories.