Research metrics can often be helpful when evaluating the impact of scholarly output, whether analyzing the influence of an individual article or the collected work of a group of researchers. Metrics have been used to support tenure and promotion, add value for grant applications, as well as to find collaborators in a particular field. Keep in mind, however, that research metrics do have some limitations. To ensure that metrics are being used responsibly, consider the following best practices.
It’s important to use more than one research metric to tell the whole story, and a good example of this is the article citation count. At first glance, if an article has been cited more frequently than others, it might be assumed that it’s a highly influential work. However, a citation is still a citation, even if a paper is cited in a negative light. This is why it’s important to use several different metrics to provide as much context as possible. Since most citation-based metrics are solely quantitative, using Altmetrics can be a helpful way to see how people are engaging with a work on social media platforms, news outlets, and citation management programs such as Mendeley.
Another best practice of using research metrics is to use a field-normalized metric, such as relative citation ratio or SNIP, when comparing different disciplines. The problem with using the h-index to compare researchers, for example, is that publishing and citation patterns are different from field to field. When examining the h-index of a researcher from biomedicine compared to a researcher from the humanities, results are likely to be very different because biomedical papers are typically cited more than in other subject areas.
To learn more about the best practices for using metrics, the Leiden Manifesto discusses the 10 principles of research evaluation, which provide great guidance. Librarians at HSLS are also available to meet with you to discuss your specific project and offer advice. You might also consider registering for the upcoming Introduction to Measuring Research Impact class at HSLS, which covers some common research metrics, when to use them, and a tour of the tools available at Pitt to help measure research impact.
If you have any questions, or would like more information about research metrics, please send an e-mail to Stephen Gabrielson at firstname.lastname@example.org.