News
How to: Measuring Usage
"Usage Statistics and Beyond“ - a workshop held on April 22-23 in Berlin
The DFG funded Open Access Statistics (OA-Statistics) project invited representatives from related initiatives, among them OpenAIRE and COAR (Confederation of Open Access Repositories). As one of those pioneering initiatives in Europe that support alternative and complementary measurements to assess the impact of research papers in Open Access repositories, OA-Statistics has build an infrastructure that gathers usage data from Open Access repositories in Germany and provides consolidated statistics.
OA statistics actively collaborates with other partner initiatives across Europe to define uniform standards for gathering, exchange and analyzes of usage data. This is a prerequisite to achieve comparable usage statistics irrespective of service-provider, repository platform, resource type or geographic location.

The presentations offered many insights on how statistics services get setup in a number of European Countries, while also highlighting technical and organizational obstacles.
- Peter Shepherd (COUNTER) introduced a new kind of measurement – the Usage Factor for Journals which is based on Journal Usage Statistics. Ross MacIntyre presented exciting results from IRUS-UK. They decided to implement a tracker code to send OpenURL key/value pairs in “push-mode” rather than the “pull-mode” based OAI-PMH approach. This enables the transfer of large amount of usage events from repositories to a central service immediately and in a very efficient way.
- Martin Fenner (PLOS-ALM) stressed that Article Level Metrics can be seen as a discovery tool that can tell data-driven stories about the post-publication reception of research papers.
- Eloy Rodrigues presented COAR's Repository Interoperability project, and a preview of the new Interest Group on “Usage Statistics” that is part of the COAR Working Group on interoperability. Jochen Schirrwagen provided results from OpenAIRE’s work-package on Impact Metrics, not to mention the collaboration of PLOS-ALM and OpenAIRE to apply Article Level Metrics on EC funded research results that have been published in PLOS journals.
The 2nd day provided room to discuss topics such as maintaining a common robot-list that can be shared across statistics service-providers for repositories and publishers, which currently hampers comparability of usage statistics. Further topics ventilated the question on sustainable business models for usage statistics service providers, the problem of gaming in social network tools that can falsify indicators in the field of altmetrics.
Last but not least, a joint interface for statistics about research papers could overcome the problem of fragmented statistics collected by numerous services. It would cover usage statistics from repositories as well as altmetrics indicators from social network tools.
The Workshop ended with the kick-off meeting of the COAR’s Interest Group “Usage Data and Beyond”, conducted by Daniel Beucke that aims to be a forum to debate on alternative measurements and indicators and to exchange experiences from statistics initiatives on an international level. The first activity of this Interest Group will be to initiate a “common used robot list to provide a global blacklist to exclude non human clicks from usage statistics”.
Links:
- OA-Statistics Workshop: http://www.dini.de/projekte/oa-statistik/english/the-project-results/workshops/usage-statistics-and-beyond
- Pirus Code of Practice: http://www.projectcounter.org/documents/pirus_cop.pdf
- IRUS-UK: http://www.irus.mimas.ac.uk
- COAR: http://www.coar-repositories.org
- OpenAIRE Usage Statistics: http://www.openaire.eu/en/support/guides/repository-managers/usage-stats
Jochen Schirrwagen (Bielefeld University, Germany)