Incentives to invest in identifiers: A cost-benefit analysis of persistent identifiers in Australian research systems
Persistent identifiers (PIDs) are unique alpha-numeric codes that positively identify entities such as people, places, and things. In addition, they are connected to registries of information about those entities, known as metadata, that enable robust linking to and between those entities. This establishes provenance and attribution, as specified by the FAIR data principles. PIDs contribute to research integrity and reproducibility by precisely identifying the resources used to conduct research and the outputs that result from it. The ability to link research activities to their inputs and outputs bolsters research integrity and facilitates the gathering of evidence for improved strategic decision-making at the individual, institutional, and national levels.
The Australian Research Data Commons (ARDC) and the Australian Access Federation (AAF) commissioned the MoreBrains Cooperative to undertake an analysis of the incentives for adoption of persistent identifiers (PIDs) by the Australian research sector. The three main benefits of PIDs are:
- Metadata reuse: PID registries act as both repositories for metadata, and as services that can provide programmatic access to it, saving the time and effort of rekeying it, and improving accuracy.
- Automation: The presence of a PID in a system or a metadata record can act as a trigger for an action. The value of automation can go beyond time saved to include more complete information and more timely information processing.
- Aggregation and analysis: At the institutional or national scale, aggregating information about entities and the relationships between them enables strategic analysis, benchmarking, the plotting of trends, and other insights.
This report sets out the benefits of PIDs primarily through the first of these lenses: metadata reuse. This is the most amenable to quantification, as data is available about the number of specific entities in the Australian research system (such as the number of researchers, institutions, publications, and grants). By combining this information with existing research on metadata use, the time taken for various kinds of manual data entry, and staff costs in Australia, we are able to attach both time and dollar values to the savings that comprehensive PID adoption could bring:
- The total time cost of this tedious work is nearly 38,000 person days per year.
- The direct financial cost of this wasted effort is nearly $24 million per year. Accounting for the opportunity cost associated with technology transfer and innovation-led growth suggests a far higher figure of $84 million per year.
Our primary recommendations are as follows:
- Develop a national PID strategy for Australia, which builds on the success of the AAF-led Australian ORCID consortium and leverages the leadership ARDC is already providing on PIDs.
- Key stakeholders in the Australian research sector—such as universities, research institutions, funders, and infrastructure providers—should integrate a suite of five priority PIDs: ORCIDs for people, ROR for institutions, RAiDs for projects, DOIs for research outputs, and DOIs for grants.
- As part of a longer-term strategy, work should continue on developing PIDs for Instruments, expanding the uses of IGSNs for samples, and potentially other IDs, in collaboration with research communities.
- Funders should build on the success of Australian Research Council’s (ARC) integration of ORCID into their research management system by adopting a similar approach and expanding to include the full suite of priority PIDs.
- Commercial providers of Research Information Management Systems (RIMS) and repositories and the communities that support open source RIMS should be engaged to encourage and enable the further wholesale adoption of PIDs into those systems.
- Ensure widespread adoption of PID workflows, with a target for 80% adoption of the five priority PIDs within five years
1 - Incentives to invest in identifiers.pdf
- Figure: 10.5281/zenodo.4991733 (DOI)