23  Retraction Watch

23.1 What it checks

The ref_retraction module cross-references the paper’s cited references against the Retraction Watch database. Citing a retracted article — usually unknowingly — is a common and avoidable problem. The module flags any cited reference that appears in the Retraction Watch database so you can check whether the citation is still appropriate.

Note

This module makes live network calls (matching references by DOI against Crossref and Retraction Watch). You need an internet connection to run the code below.

23.2 Running the module

demopaper() cites a reference that is in the Retraction Watch database.

paper <- demopaper()
mo <- module_run(paper, "ref_retraction")
mo$traffic_light

#> [1] "info"

mo$summary_text

#> [1] "You cited 1 article in the RetractionWatch database."

The table lists each cited reference found in the database:

mo$table[, c("doi", "text")] |>
  knitr::kable()

doi	text
10.1177/0956797614520714	Gino, Francesca, and Scott S. Wiltermuth. 2014. “Evil Genius? How Dishonesty Can Lead to Greater Creativity.” Psychological Science 25 (4): 973-81. https://doi.org/10.1177/0956797614520714.

23.3 Running on many papers

mo <- module_run(psychsci[1:10], "ref_retraction")
mo$summary_table

23.4 Interpreting the result

A flagged reference is not automatically a mistake — retractions happen for many reasons, and a retracted paper is sometimes cited deliberately (for example, to discuss the retraction itself). The module’s job is to make sure you are aware, so you can decide. Follow up each flagged DOI before acting.

23.5 Options

ref_retraction takes only the paper argument.

23.6 Updating the database

The module matches references against a data frame called retractionwatch that ships with Metacheck. New retractions are recorded constantly, so the bundled copy gradually goes out of date. Two functions help you manage it:

rw_date()     # the date the bundled database was last updated
rw_update()   # download the newest version from Retraction Watch

rw_update() downloads the latest Retraction Watch data (>50 MB), summarises it into a compact form (~0.5 MB), deletes the large original, and caches the result so subsequent checks use the fresh data. Run it occasionally — or before an important check — to make sure you are not missing recent retractions.

23.7 Notes

This is the same kind of check that reference managers such as Zotero perform, brought into the Metacheck workflow so it can run alongside the other reference checks.