Friday, 16 January 2015

Important lessons learnt

This blog post considers some of the lessons that we have learnt from the project to date and in particular the short pilot exercise that we undertook with 4 departments in Autumn 2014 (see earlier blog post).

Researchers and the name ambiguity problem

In talking to researchers and in responses to the pilot exercise survey, there seems to be a shared recognition of the problems of incorrect assignment of research outputs and the ambiguous nature of names. The ability to “distinguish myself” (chosen as our tagline in initial advocacy work) came top of the perceived benefits of ORCID iDs to individual researchers. There were a number of verbal anecdotes recited concerning misattribution of research outputs. Chairs of Departmental Research Committees appeared to quickly associate ORCID iDs with ideas around citation impact and research profile, both at an individual and departmental level. This is evidenced by the introductory emails that some Chairs of Departmental Research Committees sent out in association with the pilot exercise.

Registration/addition of ORCID iDs via Pure - researchers views

The implementation of ORCID iDs within Pure has consequences not necessarily relating to the perceived benefits of the IDs themselves. Pure is a central system at the University but, like other CRIS, it can be can hold negative associations for individual researchers. It is possible that implementation via a Pure route lessens the number of researchers likely to register for an iD - further investifation is required. It is noticeable that the number of iDs registered to users increased by approximately 200 during the pilot stage whilst the number of iDs listed on Pure increased by approximately 70 (see also technical aspects below). A small number of respondents to the survey suggested that getting an iD was just an additional bureaucratic burden.

There is some evidence from the pilot exercise survey that although researchers could see the value of ORCID iDs, they could not see the value of creating/adding an iD into Pure. This could be linked to the ORCID profile versus ORCID iD issues covered below but we need to be clearer on what we see as the advantages of this route. It does not seem unreasonable to spell out the advantages of having an ORCID iD in Pure to researchers. It is unfortunate that it was not possible to get a functioning EPrints connector in place (we are still working on this at time of writing) as this would be one illustration of the how the iD would link through to other systems.

ORCID iDs via PURE - technical aspects

Technical issues delayed the progress of the project at a number of stages (see also Technical approaches blog entry). This was disappointing in that we were subsequently unable to reach all of our target outcomes within the JISC-ARMA timescale. Other institutions need to be aware that all technical aspects need to be thoroughly tested ahead of any implementation to make sure that they work as intended. New functionality may have unforeseen requirements. For example, it was anticipated that the changes required to the EPrints connector would be straightforward but the ORCID iD field was initially found not to be indexed and was therefore unavailable to add into the connector (this has now been rectified).

The use of the ORCID iD functionality in Pure also proved to be more problematic for researchers than anticipated. After creating or adding an iD, the user needs to save the changes. Unfortunately, the save option at the foot of the screen was missed by many researchers. On analysis of the pilot survey responses, 10 of the 17 researchers who stated that they had created an iD in Pure had actually failed to save the iD in Pure. A similar issue appears to have applied to those researchers adding an iD into Pure. As a consequence of these issues, the University’s ORCID web page used for full implementation have additional detailed instructions on the ORCID options. A request will be made to Pure to consider a “you have unsaved changes” warning message.

ORCID profile versus ORCID iD

The York pilot exercise highlighted the need for clarity in distinguishing the ORCID profile functionality from the role of ORCID iDs as persistent identifiers - this can be a potential source of confusion. Any institution planning an implementation may wish to consider its position and targets in terms of ORCID iDs and ORCID profiles for constituent user groups.

At York, we needed to clarify our use case for academic and research staff to make it clear that for this group we are actually only interested in the iD element of ORCID. The profile element, at least from an institutional perspective, is being provided by Pure and will made public via the Pure portal (the York Research Database) in due course.

Individual researchers, however, may associate ORCID with its profile functionality, especially as they see the opportunity to add Personal information when they activate their record. This led some researchers in the pilot exercise to assume that they had to develop their ORCID profile. A significant number of researchers expressed their disappointment that Pure did not automatically populate their ORCID record; there currently no interaction between ORCID and Pure in terms of profile content other than the iD. Evidence from pilot strongly suggests that there was an expectation from researchers that one would automatically populate the other.

Postgraduate Research students and support staff - other use cases and workflows

The University of York use case is actually different for staff without a Pure profile (at York this includes support and technical staff) and postgraduate research students (who also do not have a Pure profile). For these users there may be a personal advantage in having both an ORCID iD and an ORCID profile (see above). The ORCID profile functionality provides a useful platform for these users to promote themselves and their research outputs. These factors mean that the messages communicated may need to vary depending on the planned outcomes and the user groups being targeted.

The workflows involved may also be different for these groups. For example, for postgraduate research students it may also be possible to store their ORCID iD within their student record; one option would be to investigate integrating ORCID iDs into the University’s student information system (SITS). In terms of the repository, the White Rose institutions are already investigating the use of ORCID iDs as a required field for deposit into White Rose eTheses Online. It is planned to similarly investigate the possibility of including an ORCID iD field in the direct deposit form for White Rose Research Online.

The importance of partnerships and institutional support

The York project team involved several key teams within the Information Directorate and beyond. The project was led by the Information Directorate’s Research Support team but input from the Digital York team and the Research Strategic and Policy Office (RSPO) was essential to progressing the work. At York, Pure is managed by the RSPO and without their collaboration the project would not have been possible.

From the start, the project had the full support of the Pro-Vice Chancellor for Research (PVCR). The PVCR provided a supporting letter for the JISC-ARMA bid and the launch emails for full implementation will be sent out under the PVCR’s name.

The participation of the 4 departments in the pilot exercise should also be acknowledged - the lessons learnt during the pilot exercise should prove invaluable to the full implementation. Of particular note was the importance of the departmental research administrators in certain departments - once administrators were convinced of the value of ORCID iDs they were very willing activists encouraging their researchers to register.

No comments:

Post a Comment