Garfield, E. Is citation analysis a legitimate evaluation tool?. Scientometrics 1, 359–375 (1979). https://doi.org/10.1007/BF02019306
URL: https://link.springer.com/article/10.1007/BF02019306#citeas

Is citation analysis a legitimate evaluation tool?

1. The number of negative citations is disproportionately larger than that of the positive ones, which are generally to point out incorrect results.
2. The citation counts include self-citations.

University of Michigan - Ann Arbor

On March 20, 2020, Gautret et al. published a paper on hydroxychloroquine as a COVID-19 treatment. On March 30, 2020, Alfred H.J. Kim published an opinion article criticizing its flaws. A search on May 10, 2020, showed the Gautret et al. paper had 646 citations, while Kim's critique had only 21. This large discrepancy illustrates how a heavily criticized paper can accumulate significantly more citations than the articles pointing out its flaws, even within a short timeframe.

Example of Citation Count Discrepancy Between Contradicting Papers

@EBRheum: https://twitter.com/EBRheum/status/1244713254011707398

Tweet with a link to a journal article that criticizes inappropriate results from another paper on the use of Hydroxychloroquine & other implications related to rapid reporting

@CarlosdelRio7: https://twitter.com/CarlosdelRio7/status/1248081364215308292

Twitter url to @CarlosdelRio7 who advocates for the retraction of the paper on Hydroxychloroquine (Gautret et al.)

Common criticisms for citation count as a benchmark for scientific performance

Higher the citation counts, the higher the quality, significance, and impact of a document.  

Underlying assumption of citation counts as quality indicators

Since single citation count is not convictive enough as a benchmark for scientific performance, a further NLP sentiment analysis could be helpful. Sentiment analysis could identify the emotions (positive, negative, and neutral) within text data so that we can classify all the citations in both articles to figure out the actual opinions.

Learn Before

Related

Learn After