Partially or fully (“silently”) withdrawn or retracted Center for Open Science preprints

In academia, one of the publishing models available to authors are preprints, which are non-peer-reviewed documents that typically serve as a prelude to submissions to peer-reviewed journals. Preprints can be published rapidly, following some moderation and screening, and because they can be updated if desired and when required by authors, and since they are generally not considered to be prior publications, preprints have become a popular medium of publication in recent years. Preprints, particularly those that are assigned a digital object identifier, are formal scientific documents, so they form part of the permanent knowledge base. Consequently, when preprints are withdrawn or retracted, for whatever reason, it is essential to document this event and treat such documents and authors as equally as any other document that or author who suffers a withdrawal or retraction. In addition, reasons for withdrawals or retractions should be formally noted by the preprint server, while basic bibliometric data, such as authors’ names, the preprint’s title, and the abstract, should remain intact, for posterity, while a retracted preprint needs to be labeled as such. When preprint servers opaquely scrub clean most or all identifiers from retracted (or withdrawn) preprints, and remove the accompanying file, then these are known as partial or full (“silent”) preprint retractions or withdrawals. This paper presents some cases of silently retracted preprints on the Center for Open Science preprint server.