Research data management

Retention requirements

Regardless of your choice to publish your dataset, every researcher also has a requirement to retain their research data for a number of years - this retention period changes depending on the nature of the research.

The relevant minimum durations are set out in the Western Australian University Sector Disposal Authority and are summarised in the Storing Research Data documents provided by Curtin Information Management and Archives. Below is a quick reference table to the requirements.

However, you should also check with any collaborators and funding bodies if they have any additional data preservation requirements.

Minimum data retention requirements quick reference table

Description Retention Requirement
Research data, analysis and results with outcomes that are classed as “Major”. Retain permanently.
Research data, analysis and results that are classed minor involving humans or animals that utilise high risk materials (eg. teratogens and carcinogens, ionising radiation or dangerous drugs.) Retain minimum of 50 years after date of publication, OR 50 years after conclusion of the project (whichever is later) then destroy.
Minor research data, analysis and results with outcomes that are classed as minor, but involving clinical trials. Retain minimum of 15 years after date of publication, OR 15 years after conclusion of the project (whichever is later) then destroy.
Research data, analysis and results with outcomes that are classed as minor, where the project involves children (participants under 18 years of age) Retain a minimum of 7 years after publication or project completion, OR until the subject/s have reached 25 years of age (whichever is later) then destroy.
Research data, analysis and results with outcomes that are classed as minor, not covered by other minor research classes. Retain minimum of 7 years after date of publication, OR 7 years after conclusion of the project (whichever is later) then destroy.
Research data, analysis and results relating to short-term research projects undertaken by students for assessment purposes (e.g. undergraduate degree projects) Retain a minimum of 12 months after the completion of the project, then destroy.
Large scale datasets collected for the purpose of future research opportunities. May include observational (such as radio telescope data) or machine generated data Curtin to keep for at least 12 months after last action.
Routine observational data generated by radio telescopes deemed unusable. Delete once deemed unusable.
Research data where consent for use has been withdrawn by the participant. Retain until notified of withdrawal, then destroy.

At Curtin, the Research Drive is the preferred location for meeting this retention period. Once your research is complete, copying your completed dataset to your research drive will ensure that the data is retained for the required period.

In most cases, not everything associated with the data will need to be retained. You should consider:

  • Costs associated with storage - using archiving and compression software (such as Winzip) will minimise the space required to store your data.
  • What data is considered central to the research project and how much of the data can be easily obtained elsewhere (eg. ABS statistics, shared imagery, standard test datasets).

Preservation

The concept of data preservation is similar to retention, but has some key differences.

Retention is usually a mandated requirement for researchers - it’s the task that ensures that a bare minimum of data will remain available in some format.

Preservation refers to having an active plan to ensure that when you do need to access your old data, it’s readily available and can be easily accessed and manipulated by whoever needs it. When making a plan for data preservation you should include activities such as:

  1. Transferring data from older storage formats to newer ones. This will ensure the technology required to access your data is still available.
  2. Transferring data from older file formats to newer ones.This will ensure that your data can still be opened by current software applications.
  3. Having multiple copies of the data in different locations. This will ensure that your data is not lost in an unexpected event, such as theft, floods or fire.
  4. Ensuring your data is well documented, such as making notes on software used as creating codebooks as outlined in the Documentation and description section of this guide. This will ensure that when you come back to access your data, you’ll be able to remember what it all means.

All of these types of activities will improve how reusable your data is for other researchers.




More resources