Why should we care about paper mills?

Paper mills are one of the pressing challenges in science today. The mass production of fraudulent research has changed the nature of misconduct, turning it from individual to commercialized industrial-scale business. However, paper mills are not only an integrity problem. Our knowledge on paper mills remains fragmented, often limited to certain countries, disciplines or entities. It creates the challenge of identification, as paper mills show different red flags and they adapt their tactics as soon as they are identified. Another challenge is the correction of the scientific record. The retraction gap is huge, and the question of whether the scientific record needs an audit and who is responsible for it remains largely undiscussed. The consequences reach far beyond trust in science. Papers originating from paper mills are included in systematic reviews and meta-research and can bias scientific conclusions. They may also distort results of scientometric and bibliometric studies, affecting citation, authorship and collaboration indicators. Forensic scientometrics can play a significant role, because many paper mills leave traceable patterns, for example in text, citations, authorship and collaboration. The talk will discuss how bibliometric and scientometric approaches can be suitable for the detection of inconsistencies and paper mills, and for measuring how they bias the scientific record.