The positive case for suppression: A guest post from the editor in chief of Clarivate’s Web of Science

This is an invited guest post related to news about two suppression reversals announced today by Clarivate.

The research process is rarely straightforward. There are a myriad of ways in which it can go wrong, from the inception of a hypothesis that goes on to be disproved, to failed experiments and rejected manuscripts, hopefully ending in the “happily ever after” of adding to the scholarly record through publication and worldwide dissemination… before starting all over again. Being able to build on the corpus of existing knowledge is essential for future discoveries and innovation: As Newton wrote back in 1675 “If I have seen further, it is by standing upon the shoulders of giants.”

Sadly, we know that even once published, many scientific results are not easily reproducible, and some are amended or retracted. Fraud and misconduct might be the attention-grabbing explanations for the lack of reproducibility in research, but more often than not, it is honest mistakes or making decisions with inaccurate or incomplete information that lead to errata, corrigenda or retraction of articles. Many have argued we need to be more honest about this – and to see retraction as a good thing. Correction of the version of record should be embraced, rather than avoided, and the stigma surrounding retractions should be removed.

Why should the same not apply for suppression?

Web of Science Journal Citation Reports™ (JCR) enables the research community to evaluate the world’s most impactful science and social science journals using a range of metrics, descriptive data and visualizations. It is also used by academic publishers across the globe to evaluate the impact of their journals relative to their field and promote them to the research community.

The 2020 JCR contains more than 12,000 journals from a subset of  the Web of Science Core Collection™ – those which are contained in SCIE and SSCI – and uses citation data from all of the  21,100 journals, 104,500 books and 205,000 conference proceedings indexed in the Web of Science Core Collection. At Clarivate™, we believe in the importance of maintaining the integrity of the scientific record and take our responsibility in helping to maintain this integrity extremely seriously.  This responsibility includes applying a rigorous selection procedure for the Web of Science Core Collection and analysing the citation data in JCR for anomalies. 

We believe that citation distortion is harmful to the scholarly record as it creates an inaccurate reflection of the connections between articles and their contribution to the scholarly network. It also leads to distorted metrics and rankings.

Every year since 2005, we have suppressed a small number of journals because our analysis indicates the citation data contained within the JCR does not accurately represent the journal’s contribution to scholarly communication. These journals show evidence of unusually high levels of journal self-citation or evidence of citation stacking between journals (or a “citation ring”). 

Suppressed journals stay listed on the Master Journal List, their content remains indexed in the Web of Science, and previous years’ data can still be viewed in the JCR. However, they will not receive new metrics this year. There is much focus on the fact that suppressed journals do not receive a current Journal Impact Factor (JIF), but no assessment of a journal’s value or importance should ever rely on just one metric. 

We do not assume or assign motive when investigating citation distortion because we believe that regardless of reason, these distortions create an inaccurate reflection of the connections between articles. The intention of suppression is not to punish, it is simply to ensure an accurate reflection of the citation network. 

An evolving approach 

Our suppression policy evolves as we discover new anomalies and adapt our processes. In 2020, we updated our methodology and parameters for journal self-citation. You can read our full policy here.  Changes include comparing each journal within its assigned category (or categories) – instead of to the entire JCR – to allow for subject-based differences in citation behavior. 

We realize that suppression of a journal from the JCR affects publishers, authors, editors and peer reviewers; we do not take this action lightly. We do receive formal appeals to the suppression decision on occasion, and we always ask for additional information to allow us to re-analyze and decide whether the original decision to suppress should stand. We acknowledge that in exceptional circumstances there may be compelling editorial reasons for an atypical level of journal self-citation, which fulfils a genuine scholarly purpose, and this year we have had successful formal appeals from the publishers of two journals. Current data and metrics for these journals, including the JIF, will be included in the September reload of the JCR.  

We have had successful appeals before, and it is important to note that this is never a “negotiation” or in response to pleas from the community, but a result of the publishers providing additional data and contextual information that provides a valid explanation for the observed citation anomalies. 

We continue to listen, learn and evolve, working with the research community to refine our suppressions process and policy.  Just as we believe that a retraction is not inherently negative, but a necessary tool for correcting the scholarly record, so we believe that suppressions are necessary and needed to ensure an accurate reflection of the citation network and the connections they reveal. 

Dr. Nandita Quaderi is Editor-in-Chief of the Web of Science, part of Clarivate.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].