The Duke University Medical Center Library & Archives provides access to tools and expertise in compiling and analyzing publication metrics. Research librarians can partner with you on bibliometrics and research impact projects. Based on the purpose and scope of your project, we will tailor our services to your specific needs.
The first step in getting started is to fill out our Bibliometrics / Research Impact Request Form and send it to firstname.lastname@example.org
- The time needed to complete these projects is highly variable. Please contact us as soon as possible, particularly if you have grant deadlines.
- The average time to complete an average bibliometric and research impact request is three weeks. Some may be faster, others longer. Particularly complex requests, such as those with multiple questions, authors outside of Duke, and those requiring non-standard databases take longer to develop, run, update, and analyze.
- Multiple iterations may be required to achieve the proper analysis or visualization. We will work with you to review initial results to ensure relevance.
- At any given time, we have multiple bibliometrics and research impact requests in the queue, and it may take 1-2 weeks to schedule our first consultation.
Requirements for getting started
- Librarians will work with you to determine the best data source(s) for your needs. Depending on the data source selected, we may require your assistance with data cleanup.
- Librarians will not interpret data and related visualizations for you. Rather, we will describe the data sources used and how the visualization was made; we rely on you for interpretation. We should review any Methods sections describing the analysis for accuracy.
- Given the expertise and significant time commitment involved in this work, it is our policy that librarians' contributions to most bibliometrics and research impact analyses be recognized with an acknowledgement in any publications or presentations.
- Investigators submitting requests for bibliometrics and research impact analysis as a major component of a grant-funded project should include library effort in the grant. This does not apply to using our services to help supplement information submitted with grant applications.
- We may ask your permission to show examples of this work to colleagues at Duke or other external groups to demonstrate our tools and techniques.
Use Case 1: Departmental Evaluation
Departments interested in measures of faculty productivity, datasets of faculty publications, and alerts of new faculty publications.
Deliverables may include:
- Clean datasets of faculty article citations
- Individual metrics
- H-indices (which calculate both quantity and impact of an individual's publications)
- Group metrics
- Research topic analysis based on journal type or text analysis
- Quality of journals published in
- Normalized citation scores
- Co-authorship network analysis
Use Case 2: Grant Application
Grants require evidence that the proposed research is valuable and unique and that key personnel have a successful record of accomplishment when working together.
- Network analysis of co-authorship among key personnel and of previous collaborations between institutions
- Subject analysis of current literature to explore gaps and opportunities for future research
Use Case 3: A New Working Group
New departments, team science initiatives, and other programs seek ways to demonstrate existing collaborations and opportunities for growth.
- Network analysis of co-authorship
- Identification of faculty with the potential to expand research and impact
- Benchmarking against other institutions
- Regular reports on productivity
We currently use a variety of tools for our bibliometrics and research impact projects, including:
- Scopus (Elsevier)
- Web of Science (Clarivate)
- VOS Viewer: Open source software for visualizing bibliometric networks
- Tableau: Open source software for visualizing graphs and networks
- EndNote: Creates databases of citations that can be exported to other tools for analysis
- Elements: Publication system within Scholars@Duke