Long-term availability of data associated with articles in PLOS ONE.

Area of Science:

• Bibliometrics
• Data Science
• Scholarly Communication

Background:

• Journal policies mandate Data Availability Statements (DAS) to enhance research data accessibility.
• Despite DAS, data retrieval remains challenging due to link rot and content drift, impacting long-term access.

Purpose of the Study:

• To assess the long-term availability of digital research objects linked via URLs and DOIs from DAS.
• To compare automated and manual retrieval rates of research data and identify factors influencing availability.

Main Methods:

• Extracted 8,503 URLs and DOIs from ~50,000 DAS in PLOS ONE (2014-2016).
• Employed automated and manual methods to retrieve associated research data, code, and digital objects.
• Manually validated a subset of 350 URLs and 350 DOIs for resource accessibility.

Main Results:

• Automated retrieval succeeded for 80% of resources, significantly higher than previous author-contact methods (10-40%).
• Manual checks confirmed high retrieval rates: 78% for URLs and 98% for DOIs.
• Digital Object Identifiers (DOIs) and repository sharing positively correlated with data availability.

Conclusions:

• URLs and DOIs are effective for ensuring long-term access to research data, mitigating issues like link rot.
• Including DOIs and repository links in Data Availability Statements is crucial for persistent data accessibility.
• The study supports the efficacy of persistent identifiers for maintaining the availability of digital research objects over time.