The edges carry two attributes: the multiplicity (how often two events occur in direct sequence in the items histories) and, as weight attribute for layout algorithms, the logarithm of the sum of durations between two vertices. As the case studied here shows, editorial management systems can be and are adapted to their context. Moving forward, the MDIIM continues to work with faculty members, areas, and program offices to prioritize - and expand - integrated management pedagogy and to develop bold new . In the second section of the results, we aim at tracing the order of the events in the editorial management system. Journal Peer Review and Editorial Evaluation: Cautious Innovator or Sleepy Giant? //--> All Rights Reserved. Department of Social Sciences, Faculty of Humanities and Social Sciences, Humboldt-Universitt zu Berlin, Berlin, Germany. A significant number of events (11,866, to be precise) released by editors affect actors with none specified roles. The patent as well as the digital infrastructure aim at supporting the editor in their work. We use the perspective of the infrastructure by studying the recorded events it has created as a result of actions by different actors. unfortunately, the editor dont respond about reject and accept. 117. Received 2021 Jul 26; Accepted 2021 Sep 20. This dimensionality reduction probably obfuscates some properties of the implemented process, such as if it may have been acyclic in higher dimensionality, which we cannot observe any more, limiting the potential for our investigation. Consensus decision-making or consensus process (often abbreviated to consensus) are group decision-making processes in which participants develop and decide on proposals with the aim, or requirement, of acceptance by all. Editors are responsible for making manuscript decisions based upon reviewer reports and their own reading of the manuscript. Accessibility We thank Martin Reinhart for data acquisition and consultation as well as Felicitas Hesselmann for data acquisition and feedback. Article proofs sent to author 4. Many journals now rely on editorial management systems, which are supposed to support the administration and decision making of editors, while aiming at making the process of communication faster and more transparent to both reviewers and authors. How does the infrastructure support, strengthen or restrain the editors agency for administrating the process? The event information was further enriched with year of submission, pseudonym of journal, and by (pseudonymized) data about the roles (editor, author, reviewer or none) of the person-IDs with regard to the respective manuscripts. As described above, to investigate the idealized process from the patent empirically, we constructed a simplified network from the recorded events for all 14,391 first-version manuscripts, in which the nodes represent the stages and edges are drawn between two events which follow one another. Of all 11,103 manuscripts which make it to a decision at least in one round, the first submitted version is rejected in the vast majority of the cases, whereas manuscripts which make it through the first round, stand a good chance to be accepted in the later stages, as is shown in Figure 1. This data represents a full inventory of manuscript version histories for the given years and journals, covering all submitted manuscripts whether published in the end, or not. To identify important passage points in the network, we chose node degree centrality with respect to edge multiplicity. Before the decision, basically two things can happen (see Figure 5). Professional Empirically, we use digital traces from an editorial management system in order to gain insights into how the digitalized peer review process looks like. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Interestingly, when Potential Referees Decline (N = 7,743), this event is mostly triggered by a none role, because declining referees do not have a role with the manuscript in question. Histograms of sums of durations between successive events in the process: The distribution is skew to the left; the log-scaled distribution is better leveled (Remark: 14 durations of length 0 are left out in the logarithmized plot). While the potential exploitation of these process generated data may support the administration, it at the same time may also put more pressure on the editor, simply because these data can be shared and discussed with potential stakeholders of the publisher. The following decision types are available: Reject; Major revision; Minor revision; Accept; Decisions are communicated to the corresponding author in a formal letter, along with reviewer feedback and any other requirements from the . Among the leading intellectuals of his time, Franklin was one of the Founding Fathers of the United States, a drafter and signer of the United States . The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Nature. That is why it would be difficult to make claims about changes between a pre-digital and a digital scholarly journal world: we simply do not know enough about organizational practices of peer review as such, though research about peer review has grown recently (Batagelj et al., 2017). Before Of major relevance for the peer review process is that it finally comes to a decision, based on consultation with internal and external actors. Given the administrative responsibilities of the editors, it is plausible that some of these events refer to quality or process control related activities such as setting up automated mailings without a call for action. a cover letter that provides any additional information requested by the editors. We have also gained specific insights into how editors take their role in the peer review process seriously: despite automation of some administrative steps, decision-making as well as decision-communication remains in the human domain. The editorial management system makes these different roles visible, by attributing person-IDs as authors, editors and reviewers to manuscripts. This network turned out to be relatively complex with 72 nodes and 623 edges, and relatively dense (with d = 0.12), which means, that 12 percent of all theoretically possible edges occur empirically. Editor in Chief, Nature. In the second category, which Schendzielorz and Reinhart (2020) have called consultation, we subsumed nine events, which are mainly performed by editors, reviewers and none roles. This category is comprised of Waiting for Editor Assignment (N = 14,261), Waiting for Potential Referee Assignment (N = 12,976), Waiting to Send Decision to Author (N = 5,796), Waiting for Revision (N = 2,612), Waiting for Author Approval of Converted Files (N = 898) and Potential Referees Waited too Long (N = 610). 1 January 6, 1705] - April 17, 1790) was an American polymath who was active as a writer, scientist, inventor, statesman, diplomat, printer, publisher, forger and political philosopher. Sometimes they are more busy. The graphic below shows how a typical manuscript goes through the Editorial Manager system, along with some of the terms used to describe the manuscript's status. According to Guston (2001), there is a social contract granting autonomy and self-regulation to science only if scientific quality and productivity is ensured. Digital infrastructures such as editorial management systems allow for processing data about the submission, evaluation and decision of manuscripts in novel ways, taking particularly the velocity, role specificity and consistence of the peer review process into account. We concentrate on the core process now and delete the now isolated vertices, thus reducing the core process to the main component of the network with 48 vertices and a density of d = 0.04. The quantitative analyses were performed with the use of R (R Core Team, 2020) and the following contributed packages: igraph (Csardi and Nepusz, 2006), tidyverse (Wickham et al., 2019), lubridate (Grolemund and Wickham, 2011), data.table (Dowle and Srinivasan, 2021) and ggraph (Pedersen, 2021). This to be acknowledged, Seaver (2017) described some tactics for the ethnography of algorithmic systems, of which we take up the tactic of scavenging in our work: using the pieces of information accessible to us while at the same time keeping in mind that we only see a part of the whole picture. 1124. The patent shows the components like postulation, consultation and decision as elements relatively clearly, but the component of administration is distributed over the whole process. The rejected manuscripts and those to be resubmitted get a special treatment by the editors: the communication about the frustrating decision is thoroughly crafted showing in the network as two vertices about Drafting Decision Letter, notably resulting in longer durations for decisions to be sent to authors. Like other journals in the Nature family, Nature Microbiology has no external editorial board involved in editorial decision-making. Moreover, the characteristics of both reviewers and editors are explored to a significant extent (Hirschauer, 2010, 73). That is why we also focus our structural analysis of the peer review process on this first round of peer review. Additionally, source and target vertices were inserted to make start and end of the process visible in plots. As Horbach and Halffman (2020, p.4) have argued, such infrastructural systems of classification and standards constitute invisible mediators of action establishing templates () by which performances are compared and which define what one enactment is a performance of (ibid). Katharina is a communications expert, science communicator, non-fictional book author and now Communications Director at the foundation "Gesunde Erde - Gesunde Menschen".<br><br>While earning her doctorate, she taught with a focus on cultural and media studies at the LMU Munich. When we plot the network with Kamada-Kawai layout, the high network density causes the network to appear as a circle (see Figure 4, left) with no visually detectable pattern between source and target. It has core editorial offices across the United States, continental Europe, and Asia under the international scientific publishing company Springer Nature. Some authors ask the editors to reconsider a rejection decision. The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frma.2021.747562/full#supplementary-material, National Library of Medicine 8600 Rockville Pike Marres (2017) points out that by dealing with data from digital infrastructures, research agency is twisted: the data often prompt the researcher to their perspective and methodology, resulting in that digital research requires an at once critical and creative approach to method (p.115). Also, Manuscript Transferred (N = 995), Manuscript Ready for Publication (N = 1,705) and Manuscript Sent To Production (N = 1,694) are events covering the transfer of publications after the review process was completed, referring to their relationship with the publishing house and their facilities. The publisher uses the system EJournalPress to manage their editorial peer review lead by full-time staff editors in a shared office space. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). We found multiple observations for each manuscript with a stage name, a time stamp and two pseudonymized person-identity numbers (hereinafter, person-IDs), in the system originally identifying individual users assigned to it the person who triggered an event and the person affected by an event (judging by the xml-tags assigned to the information). Yet, in our data set, we also found events that reach beyond administrative activities, because they document pace, effectiveness, or quality of the process or the item (the manuscript), thus enabling quality control and supervision of the whole process, which we label observational elements. So to reduce the noise and to uncover the core process, we deleted all edges, which had a multiplicity of less than 1% of the number of items. Due to the specific work environment at the publisher, where editors are employed as full-time staff in a shared office space, it must be easy for them to communicate with each other bypassing the editorial management system, which limits the potential of surveillance through the system. Comparisons with novel digital infrastructures (and their implementations) for other publishers with different peer review models are necessary in order to more systematically judge or reflect on the influence of these infrastructural tools on innovation or stabilization in editorial work. In the minimal process of peer review according to Schendzielorz and Reinhart (2020), we would find the four processual elements being mutually connected with each other.