Hutchins, B. I., Yuan, X., Anderson, J. M., & Santangelo, G. M. (2016).

Relative Citation Ratio (RCR): A New Metric That Uses Citation Rates to Measure Influence at the Article Level.
PLoS Biology, 14(9), e1002541. https://doi.org/10.1371/journal.pbio.1002541

Review by Tamas Nagy, iPQB Graduate Program, Weiner Lab, tamas@tamasnagy.com

This paper introduces a new bibliometric algorithm for measuring the influence of scientific articles in the biology field. Many bibliometric algorithms have been proposed that compare the citation accumulation rate to an expected citation rate. The primary difficulty and where approaches differ is on how to calculate the latter value. The main breakthrough of this paper is to compute an expected citation rate for a research article (RA) by looking at the performance of the peers of RA. The authors define peers as being papers that co-occur with RA in the references list in articles citing RA. They show that these articles are much more similar to RA than articles published in the same journal. They go on and compute an expected citation rate by regressing the citation accumulation rate of a large set of papers onto the citation accumulation rate of the journals their peers are published in. Each article then gets assigned a relative citation rate (RCR) where its actual citation rate is divided by the expected citation rate. This approach benefits from being resistant to citation style differences between fields and to gaming via self-citation. They also show that their metric stabilizes relatively quickly so it does not disadvantage younger scientists. These advantages are not shared by its competitors and could be why the RCR is being adopted by major funding agencies including the NIH, NIGMS, and the Wellcome Trust. Weaknesses of this paper include that it did not directly compare its algorithm versus more recently published bibliometric algorithms (other than h-index and Journal Impact Factor). Additionally, I was not convinced that using the average journal citation rates for computing the expected citation rate was the best approach and the authors do not show data on the performance of this approach versus simply taking the median citation accumulation rate of a RA’s peers as the expected. Lastly, the data used only came from Pubmed so papers with interdisciplinary references can be hurt in this metric.

I am studying the regulatory networks and dynamics of cellular volume homeostasis. I am broadly interested in systems biology, bioinformatics, and microscopy.