Is the soundness-only quality control policy of open access mega journals linked to a higher rate of published errors?

Abstract

Open access mega journals (OAMJs) are broad and centralized open access journals that have come to represent profitable outlets for accumulating large volumes of research from multiple fields of study, including papers that are rejected from other journals by the same publisher. Some OAMJs charge hefty (exceeding US$1000) article processing fees. One characteristic of OAMJs is a large editorial board. In 2015, Björk indicated that a primary characteristic of an OAMJ was its prepublication soundness only peer review, i.e., novelty, significance, relevance and impact are assessed only post-publication. However, such a premise ignores the inherent nature of peers’ bias. This controversial claim is challenged in this paper by assessing whether there is a link between research output (number of papers published in each OAMJ) and number of errata, including retractions. We assessed 16 OAMJs indexed in Clarivate Analytics’ Web of Science and found wide variation in published errata and retractions across OAMJs between 2012 and 2018. PLOS ONE had the highest correction rate (3.159%), followed by Medicine (3.158%), BMJ Open (2.949%) and Scientific Reports (2.896%). In contrast, PeerJ, Elementa, and Science of the Anthropocene did not publish any errata in 2012–2018 but IEEE Access had a correction rate of 0.059%. Regarding the retraction rate, the highest share of retracted publications was seen in Medicine (0.079%), Cell Reports (0.035%) and PLOS ONE (0.030%), while nine out of the 16 studied OAMJs did not have any retracted publications during 2012–2018. We conclude that there is wide variation in “quality control”, as assessed through errata and retractions, among OAMJs. We recommend, therefore, that the “soundness only peer review” prerequisite for OAMJs should be scrapped.