Detecting anomalous referencing patterns in PubMed papers suggestive of author-centric reference list manipulation

Abstract Although citations are used as a quantifiable, objective metric of academic influence, references could be added to paper solely inflate the perceived influence body research. This reference list manipulation (RLM) take place during peer-review process, or prior it. Surveys have estimated how many people may been affected by coercive RLM at one time another, but it is not known authors engage in RLM, nor what degree. By examining subset active, highly published (n = 20,803) PubMed, we find frequency non-self-citations (NSC) author coming from single approximates Zipf’s law. Author-centric deviations approximately normally distributed, permitting quantified statistically. Framed an anomaly detection problem, statistical confidence increases when outlier multiple metrics. Anomalies proof engaged will almost unavoidably create anomalies. We NSC Gini Index correlates with anomalous patterns across “red flags”, each suggestive RLM. Between 81 (0.4%, FDR < 0.05) and 231 (1.1%, 0.10) outliers on curve, chronic, repeated Approximately 16% all some Authors who use 18% more their for self-citation significantly likely distortions, suggesting potential willingness coerce others cite them.