Purpose & Goals

Various campus entities use bibliographic data to 1) analyze and benchmark the institution’s position in relation to other universities, 2) understand collaborative networks, and 3) explore interdisciplinary connections. This data, however, is frequently not used in coordinated ways. Units independently harvest, repackage, and report it using their own procedures, definitions, and practices, leading to inconsistent results. Academic librarians are uniquely positioned to address this issue. As trainers and metadata experts, they already educate campus constituents to locate, gather, utilize, understand, and apply bibliographic tools and measures to assess research impact, both ethically and effectively. But stretched campus institutional research and research analysis units still invest a tremendous amount of time and human capital matching data collected from various public and proprietary sources to local discipline groupings, research programs, and other analysis units. Academic libraries can help these groups save time by using their data skills in concert with existing campus infrastructure to help facilitate consistent, timely reporting. This study asks how academic libraries can engage campus analysts and use existing taxonomies with unique identifiers to develop tools that save campus planners time and improve reporting consistency across units. The study offers new ways for envisioning the application of library assessment by showing how integrated library research impact and data visualization services can successfully position the library as a partner for campus organizational performance improvement while simultaneously helping researchers make scholarly work discoverable and administrators understand the breadth, scope, and success of the university’s various research programs.

Design & Methodology

Aware of an ongoing campus need to link funding data to publications, patents, and clinical trials, and filter this data using local taxonomies, the authors developed a list of use cases to share with campus planners to surface mutual goals and guide the development of tools. Because the possibilities of this type of work are infinite, the authors believed a prototype was needed to inspire discussion and focus the linking of bibliographic data to local systems. They initially developed a series of Tableau dashboards that integrate NIH RePORTER, iCites, and PubMed data. These dashboards not only raise awareness of the existence of and richness of the NIH data but show how to meaningfully model and visualize both grant and bibliographic data, as well as filter this data using local constructs. The prototype includes federal funding data for grants awarded to the university between 2018-2022, lists of research publications associated with this data from the NIH RePORTER, corresponding reference and citation lists along with RCR values from iCites, and bibliographic information for the reference and citation lists from PubMed. To allow meaningful filtering, the data is linked to local systems using a simple join table that matches local identification numbers for each author with the research publication PMID, and the Scopus author id. Since most campus researchers do not have the requisite skills to harvest and clean author and location data from the messy affiliation field in PubMed, the join table helps campus planners to normalize faculty names, as well as university college, department, and division assignments. They can filter data using local directory information - including faculty ranks, discipline groupings, research programs, buildings, and more.

Conclusions

The prototype is now complete, and the authors are engaging various constituents across campus using the use cases and prototype. The dashboards include embedded instructions for using the views and document the sources the data was assembled from. A data dictionary is also available to facilitate a shared understanding of how each field is defined. When ready, the authors plan to mount the join table to Tableau Server and start educating campus partners on how to apply this table to their own projects to consistently filter data using meaningful local taxonomies. We anticipate additional questions concerning how research impact and data visualization services can broadly support organizational performance measurement will result from these meetings. We will also work with partners to develop future tables linking local data to other bibliographic data sources to facilitate benchmarking and comparisons.

Implications & Value

As research impact and data visualization services evolve, librarians in these roles have an opportunity to work with campus partners to support the university’s organizational performance measurement. This study builds the library’s existing reputation for providing quality bibliographic research and instruction with the creation of new services that leverage librarian expertise to save researchers time and improve the quality of their analyses. It also provides an example of the creation of a cost-effective data source by harvesting public data and combining it with local datasets which can be replicated by other institutions.