Understand the range of institutional system types and approaches
A key aspect of ORCID implementation is to determine
- in which systems ORCIDs need to be captured; i.e. the systems that should be used to present researchers with the opportunity to register or to connect to an existing ORCID
- to which systems the ORCIDs need to be propagated so they can be stored
Linked to these decisions is whether one system should be regarded as the canonical source by all other systems; that would typically be the single place where ORCIDs are captured.
In terms of capture, there is an argument that institutions should choose a single canonical source (aka ‘source of truth’) from which ORCIDs and other key person data can be propagated systematically and reliably across their other ‘downstream’ systems.
The downside of that strategy is that, with the possible exception of a Directory Service, there is no system that wants to know about everyone who might want to or be encouraged to register for an ORCID; for example:
- CRIS: only for researchers and perhaps only for researchers with grants
- Institutional Repository: perhaps open to all but not a place to register people without a deposit (for example, new postgraduates)
- HR: only likely to contain staff, and perhaps only staff with employment contracts, therefore missing those such as visiting researchers
Whilst these restrictions do not matter if the university is only targeting ORCID adoption amongst recognised ‘researchers’, there are issues for those with a broader engagement strategy. Some universities have therefore adopted a ‘hub’ approach that isolates the job of capturing ORCIDs , and then propagates them across all the systems requiring ORCID links (e.g. this was the plan for University of Southampton in their original pilot project in 2015 https://eprints.soton.ac.uk/373527/ although their strategy later changed. A more recent consideration and implementation of the approach is reported by the University of Oxford http://blogs.bodleian.ox.ac.uk/digital/2016/07/29/orcid-scoping-study-report/) [Updated 18/07/2018]
There are two approaches to propagation of ORCID IDs to ‘downstream’ systems, the choice being dependent on the capabilities of the system.
- EITHER enable downstream systems to validate ORCIDs from your source system whenever they are entered, perhaps pulling in the ORCID as soon as the user enters their name. This type of capability is dependent on how the system vendor or the OS development community has integrated ORCID.
- OR simply enable entry of an ORCID in appropriate field without validation. This approach is prone to error and probably discouraging to the user but easy to implement, except in systems with no fields that could be designated for such use.
In both cases, bear in mind that inclusion of an ORCID in a system (such as a Repository or the HR system) can be much more powerful, in terms of data interactions, where the university has enabled the individuals to connect their ORCID with the university – otherwise ORCIDs are just unique but relatively dumb IDs.
The underlying recommendation is that selecting a single system where the core research community can register for new ORCIDs and affiliate current ORCIDs is a minimum position if the university and the researchers are to benefit from real integration through the ORCID APIs.
Your selection of that system will be dependent on its ORCID integration capabilities as well as on your preferred strategy based on the principles set out here.
The options you might consider are set out in the following table, which is populated with a fictitious scenario.
|Role||Capture ORCIDs and integrate with ORCID services||Use validated ORCIDs through integration||Enter ORCIDs in systems without validation|
|Grant Management||(Possibly at a later date)|
|Research Publications Repository||Yes|
|Research Data Repository||Yes|
|HR / Personnel||Yes|
|Hub system for ORCID capture||n/a||n/a|
Examples of institutions who use their CRIS for registration are
- Imperial College using Symplectic: https://www.imperial.ac.uk/research-and-innovation/support-for-staff/scholarly-communication/orcid/create-an-orcid-id-and-add-it-to-symplectic/
- University of York using PURE: http://www.york.ac.uk/library/info-for/researchers/orcid/create-in-pure/
- ANDS, the Australian National Data Service, has provided an informative introduction to levels of implementation and integration: http://ands.org.au/__data/assets/pdf_file/0005/415535/orcid-implementation-options20150414.pdf