“Five steps to decide what data to keep: a checklist for appraising research data”

Themes

“Five steps to decide what data to keep: a checklist for appraising research data”

What is this about?

One of the Data management practices is keeping and storing research data. Since it is not always easy to decide what data should be kept for future use, researchers can benefit from consulting relevant practical tools that could guide them through their decision-making process. One of these guides is a checklist called “Five steps to decide what data to keep” that proposes five steps to follow in the process of selecting data to keep and store for long-term.

Why is this important?

According to the revised edition of the European Code of Conduct (ECoC), researchers and research institutions should “ensure appropriate stewardship and curation of all data and research materials, including unpublished ones, with secure preservation for a reasonable period”. Keeping research data is beneficial on many different levels – it is important for replication, checking validity and proving intellectual property, and also for preventing questionable research practices or research misconduct, such as plagiarism.

During the entire research lifecycle researchers make choices about data they use. As they prepare for research, they search for relevant data and select them from various sources. Then they choose data suitable for analysis. In the end, they select and chose analysed data to inform about the research findings. Since not all data should be preserved for long-term, it is important to select relevant data to keep and store in repositories. Therefore, researchers should be able to make informed decisions about what kind of data to keep to meet their research objectives.

For whom is this important?

ResearchersResearch institutionsPhD studentsLibraries

What are the best practices?

The purpose of “Five steps to decide what data to keep” checklist for appraising research data, developed by the Digital Curation Centre (DCC) is to guide researchers and help them choose what data to keep in order to meet their objectives and comply with the institutional and external funders demands. These are five steps proposed by the Guide:

Step 1. Identify purposes that the data could fulfil

Before deciding what data to keep, researchers should be able to identify goals and objectives of these data: verification, further analysis, building academic reputation, community resource development, further publications, learning and teaching, or private use. Identifying one of more of these purposes will make the selection process go smoother and faster.

Step 2. Identify data that must be kept

This concerns legal principles that should be respected. With regard to this issue, researchers should answer following questions: Are there Research Data Policy reasons to keep it? Do regulations require the availability of the data? Are there other legal or contractual reasons? Does it contain personal data relevant to the reuse purpose?

Step 3. Identify data that should be kept

This step concerns researchers’ expertise and knowledge on research data. Before making a decision regarding this aspect, they should consider following questions: Is it good enough? Is there likely to be a demand? How difficult is it to replicate? Do any barriers to further use exist? Is it the only copy?

Step 4. Weigh up the costs

Considering this step will help researchers in economic aspects of data keeping, i.e. potential costs of all the stages of data keeping - creation, collection & cleaning, short-term storage & backup, short-term access & security, team communication & development, preservation & long term access, and staff time.

Step 5. Complete the data appraisal

This final step includes filling out the table on data collection, reuse purposes, value, risk of budget shortfall and whether you plan to keep your data.

Andrijana Perković Paloš contributed to this theme. Latest contribution was Oct 18, 2021

Other information

Who

• Digital Curation Centre

Virtues & Values

• Transparency
• Availability

Good Practices & Misconduct

• Data management
• Open access