key: cord-0618919-o0rpkt63 authors: Rousi, Antti Mikael; Laakso, Mikael title: Overlay journals: a study of the current landscape date: 2022-04-07 journal: nan DOI: nan sha: d61f41b0d97d21f2132d78017e67a7bc6344fd52 doc_id: 618919 cord_uid: o0rpkt63 Overlay journals are characterised by their articles being archived on public open access repositories, often already starting in their initial preprint form as a prerequisite for submission to the journal prior to initiating the peer-review process. In this study we aimed to identify currently active overlay journals and examine their characteristics. We utilised an explorative web search and contacted key service providers for additional information. The final sample consisted of 35 active overlay journals. While the results show an increase in the number of overlay journals in recent years, the current presence of overlay journals is diminutive compared to the overall number of open access journals. The majority of overlay journals publish articles in the natural sciences, mathematics or computer sciences. Overlay journals are commonly published by groups of scientists rather than formal organisations and overlay journals may also rank highly within the traditional journal citation metrics. Nearly none of the investigated journals charge fees from authors, which is likely related to the cost-effectiveness of the overlay publishing model. Both the growth in adoption of open access preprint repositories, and researchers willingness to publish in overlay journals will determine the models wider impact on scholarly publishing. Digitization of the scholarly publishing industry in the 1990s enabled major efficiency gains compared to the operating circumstances dictated by the paper-based past. Large publishers were able to benefit from even larger economies of scale both in production and sale of read-access to published materials, and for smaller actors the cost and technical competency required to run a journal have become lower as open source solutions like Open Journals Systems have matured. However, the functional logic and service offering of journals have remained largely the same despite the shift to digital. So far, growth in open access (OA) through journals has largely been driven by outlets that adhere to manuscript handling, peer-review, and publication processes that are not radically different from what print-based processes enabled or required. Alternative publishing models that more fundamentally leverage the circumstances of digital-only and open access have been evolving, and among them is the overlay journal model (Herman et al., 2020) which is the core focus of this study. Overlay journals is an emerging model within academic journal publishing, but conceptually and practically it is nothing new as it has already been brewing in relative obscurity for the last two decades when it comes to widespread uptake (Smith, 2000; Pinfield, 2009; Thornton & Kroeker, 2021) . Prior literature has also referred to such journals as deconstructed journals or superjournals (Smith, 1999; Eysenbach, 2019) . Overlay journals mix together elements from the gold and green routes to open access which have often been presented as separate paths to enable open access. In the context of OA, the gold route is commonly depicted as publisher-provided OA where the final published articles are freely available through the journal's website to any interested reader (Laakso & Björk, 2013) . The green route is author-provided OA where the manuscript version of the work is deposited into an OA repository or to the researcher's personal website from which it is freely available to any interested reader (ibid.) . With the overlay model the gold OA element comes from there being a journal that curates content through editorial work, manages its peer-review, and ultimately makes the final output available OA without any paywall for readers. The green element comes from overlay journals basing their manuscript handling on public OA repositories which function as the cornerstone of overlay journals (see e.g. Pinfield, 2009 ), repositories such as arXiv, biorXiv and HAL. As is reviewed in closer detail as part of the study in this paper, the processes of submission/materials acquisition and peerreview management differ among overlay journals but in the most common form authors first upload their preprint to an OA repository and the link to the manuscript is sent to the editors of the journal as part of the submission processes. After peer-review and formal acceptance, which is managed by the journal editors, the authors upload a new version of the manuscript to the OA repository and the overlay journal links to the final version of the work and provides journal issue and volume information for it (hence "overlay"). The growth of preprints and the recent increased momentum around overlay journals can be argued to be interconnected. Similarly to overlay journals, the practice of posting preprint versions of journal article manuscripts is nothing new and has been around for several decades even before the world wide web, but in particular the COVID-19 pandemic has fuelled new growth and acceleration into the discussion about the topic concerning a broad spectrum of research disciplines (Fraser et al., 2021) . A challenge that the growth of preprint distribution has been facing is compatibility with the journal publishing sector, where many journals have unclear policies on permissibility or restrict such practices for manuscripts submitted to the journal or (Klebel et al., 2020) . The connection between preprints and journal submission and publication have been studied with results suggesting diversity in how and when authors upload preprints, but that most uploads seem to happen around the time that authors submit their manuscript to a journal (Larivière et al., 2014; Anderson, 2020) . The overlay journal model can potentially enable several benefits in comparison to traditional ways of publishing and communicating research results. As mentioned in the previous paragraph, overlay journals are highly compatible with and build upon the growing practice of researchers making preprints of their manuscripts available prior to formal peer-review of the content, thus potentially bringing more speed and openness to the research communication processes compared to publication of formal journal outputs. The use of existing digital repositories and their technology has been seen to reduce publishing costs related to, e.g., content acquisition, dissemination, and archiving for overlay journals (Grossmann & Brembs, 2021, p. 3 ). The overlay model may thus be seen as a cost-effective model of open access publishing that "piggy-backs" on the technology already provided by the digital repositories (Grossmann & Brembs, 2021, p. 10) . For self-managed journals the issue of preservation of published materials is not something that happens automatically and some journals have vanished together with their published materials (Laakso, Matthias & Jahn, 2021) . Preservation of published materials is also likely managed well through the overlay publication model as existing repositories with robust backups and preservation services are utilised. The key research articles looking into overlay journals are at this point over a decade old (Smith, 2000; Pinfield, 2009) , which is a long time considering the pace of change in this space. An exception to this is a recent conference paper by Thornton and Kroeker (2021) where the authors summarise the state of and growth of the largest subject repositories and compiled a list of past and presently active overlay journals together with their key characteristics. The present work extends upon this foundation by focusing on currently active overlay journals, analysing their characteristics in depth including manuscript source repositories, potential article processing charge for authors, research disciplines, publisher types, content licencing, publication volume, material submission and peer-review processes, open data, as well as key indexation and citation metrics. Although the definition of an overlay journal may vary, this study utilises the following criteria: for peerreviewed original articles, the journal 1) mandates the submission of pre-prints to an open access repository, and/or 2) hosts the final published articles in an open access repository. The research aim of this article is to identify all journals currently operating with an overlay model and to examine their characteristics. The main research questions are posed as follows: • RQ1 What journals are currently operating with the overlay model, and what are their fields of research? Once the current overlay journals are identified, the present study seeks to analyse their characteristics as follows: • RQ2.1 When were the journals started and when did they start publishing based on an overlay model? As no definitive list of overlay journals exists, the identification of the sample journals was an explorative process. The sample of investigated overlay journals were prevalently identified by using the websites of Episciences.org (2021), Scholastica (2021) , Free Journal Network (2021), Open Journals (2021), PubPub, and Wikipedia (2021) . The Scholastica and PubPub teams kindly provided information of their overlay journals to the authors. In addition, Twitter (2021) was searched with terms such as "arxiv overlay" and "overlay journal" to identify additional journals. The sample of journals for this article totalled 35 active overlay journals currently accepting submissions. The identification of sample journals was done during April 2021 to February 2022. Platform-based venues, such as F1000 Research and Open Research Europe, were not examined in this study as they do not utilise an external repository as part of the submission or publication process. In addition, journals publishing software that refer to GitHub for source code, such as Journal of Open Source Software, were not treated as overlay journals in the present study. Furthermore, rapid review and annotation type of venues, such as Rapid Reviews Covid-19 were excluded from outlets included in this study. Although these venues utilise open access repositories and facilitate peer-review of preprints, they do not incorporate a structured peer-review process that would lead to an authoritative editorial decision for inclusion/exclusion of materials in the venue. In addition, journals that previously operated using the overlay model, but either are no longer active, i.e., do not accept manuscript submissions (e.g. BiOverlay), or have transferred to different publishing models (e.g. Journal of High Energy Physics), were excluded from the study. The journal ISSN numbers, manuscript source repositories, first overlay volumes, article volumes (if not available in Scopus), publication languages, peer-review type, licence for published articles, author costs, publisher types, submission policy, and pre-print availability policy were observed by inspecting journal editorial policies and submission guidelines found from journal websites. The overlay journals' ISSN numbers were identified by examining journal websites and cross-checking this information with the Ulrich's periodicals database (Ulrichsweb, 2021) . Journals that published review reports, either with reviewers' names or anonymously, were classified as to operate with open peer-review. Publisher types defined by Laakso and Björk (2013) were used to categorise the findings concerning the publishers. If the journal website did not include publisher information, the editorial board was interpreted to publish the journal. As the sample journals were not comprehensively covered by any catalogue providing field of science information (e.g. Ulrichsweb database or DOAJ), the Organisation for Economic Co-operation and Development (OECD) (2007) field of science classification was used to categorise the journals into different domains of science. The journals' primary OECD field of sciences were defined by the authors through examining the journal websites. Exceptionally, the ST-Open and Journal of Interdisciplinary Methodologies and Issues in Science journals were given a multidisciplinary classification, which as a category is not listed in the OECD field of science classification. Whether the journals were indexed in the Directory of Open Access Journals (DOAJ), Scopus, or Clarivate Analytics' Web of Science Core collection's journal master list was examined by searching the services with journal ISSN numbers and journal titles (DOAJ, 2021; Elsevier, 2021; Clarivate, 2021a) . If a journal was found to be indexed in the Scopus database, the article volumes for the years of 2018-2021 were extracted. For journals and volumes not indexed in Scopus article counts were manually counted from the journal webpages. The identified overlay journals were examined from the viewpoint of both qualitative and quantitative journal metrics. The qualitative metrics comprised the Nordic expert panel rankings of scientific journals, namely the Finnish Publication Forum (FPF) (Finnish Federation of Learned Societies 2018), the Danish Bibliometric Research Indicator (DBRI) (Ministry of Higher Education and Science, 2022) and the Norwegian Register for Scientific Journals, Series and Publishers (NRSJP) (Norwegian Directorate for Higher Education and Skills, 2022). The Finnish Publication Forum and the Danish Bibliometric Research Indicator classifications place journals into one of three levels: 1 = basic; 2 = leading and 3 = top. The Norwegian Register for Scientific Journals, Series and Publisher operates with two levels of classification: 1 = basic and 2 = leading. It is noteworthy that these classifications include journals by suggestions from the public, so exclusion from the listed journals does not mean that the journals could not be considered as scientific. Searches were conducted from the web portals of the above services with both ISSN numbers and journal titles. To add a quantitative citation metric into the investigation, Clarivate Analytics' Journal Citation Reports database (Clarivate, 2021b) was searched with the use of both ISSN numbers and journal titles to identify whether the journals had a Journal Citation Indicator (JCI), Two-Year Impact Factor (IF) and an Impact Factor ranking (IF rank). The examined Journal Impact Factors and Impact Factor rankings were for the year 2019 (as released in 2020). The collected data are provided as open data to facilitate future research. The data are openly archived here https://doi.org/10.5281/zenodo.6420517. This study was able to identify 35 active overlay journals. From the 35 journals examined, a total of 30 (86%) had an ISSN number. Table 1 presents a listing of the identified journals categorised by research field. Table 1 Journal titles, ISSN numbers, and URLs sorted into research field categories. The two research areas with the most journals were computer and information sciences (n=9) and mathematics (n=9). This was followed by physical sciences (n=4) and basic medicine (n=2), biological sciences (n=2), economics and business (n=2), multidisciplinary (n=2), clinical medicine (n=1), mechanical engineering (n=1), philosophy, ethics and religion (n=1), and psychology (n=1). Since journal sizes can vary a lot, we wanted to gain a better understanding for the differences in the different fields of research by investigating the number of published articles in the journals during the four-year period of 2018-2021. Figure 1 presents the results of this analysis. With a total of four overlay journals the physical sciences journals output the most articles (326 in total) in 2021 with an increasing trend over the last few years, while the eight journals in computer and information sciences came in at second place for 2021 with a decreasing trend of article output (206 in total). Eight mathematics journals provided a stable total output of around 50 articles for each of the four years. The previous research question already alluded to the fact that overlay journals have not necessarily adopted the model from their inception but have rather converted to it at some point in their lifecycle. Figure 2 visualises the journal start years and conversion years to an overlay model. Some journals such as the Hardy Ramanujan Journal have retrospectively uploaded articles to an open access repository after making the transfer to the overlay model. The results indicate a broad mix of journal backgrounds and research fields contributing to the current composition of outlets in the landscape. Notable is that almost half of the journals (n=17, 49%) started as overlay journals from 2020 or later. We found that the investigated overlay journals did in general not charge fees either from readers or authors. Only one recently converted overlay journal, eLife, charged author fees and The Open Journal for Quantum Science had a voluntary article processing charge for the authors. 30 (86%) of the investigated journals explicitly stated having a no-fee policy. A total of three (9%) journals did not provide information about author charges on their websites, which we assumed to mean that no payment would be required. A clear majority of the investigated journals operated using peer-review that remained blind throughout the entire process (n=29; 83%). Only two (6%) of the investigated journals reported to conduct open peer-review by publishing reviewer reports. In addition, two (6%) of the journals accepted both blind and open peer-reviews depending on the preference of authors. Two of the investigated journals did not explicitly define their peer-review process or policy. All the investigated overlay operated with the principle of open submissions, i.e., none of the journals operated with only invited submissions. Twenty-six (74%) of the investigated journals only accepted submissions of preprints openly available in repositories prior to being sent to the journal. Although 9 of the overlay journals also accepted submissions of manuscripts that were not available in preprint servers, most often these journals also encouraged authors to take the step of uploading their manuscript to a preprint server prior submission. All but one (97%) of the investigated overlay journals hosted the final peer-reviewed articles only or also on preprint servers. eLife was the only journal that did not currently host the final articles in a preprint repository. However, eLife's new policy mandates the availability of preprints for peer-reviewed articles and thus was included in the list of investigated journals. Twenty-five (71%) of the investigated journals both mandated the open availability of submitted preprints and hosted the final articles only or also in a preprint repository. arXiv and HAL were the most frequent repositories utilised by the investigated overlay journals. Twentysix (74%) of the investigated journals allowed submissions from arXiv and 17 (49%) of the journals allowed manuscript submissions from HAL. Figure 3 presents the primary field of research of the investigated overlay journals and the distribution of which repositories are used for enabling the overlay model of operation. Figure 3 Repository use of overlay journals, sorted into journal discipline categories where each row represents one journal. Although the investigated overlay journals most often represented computer and information sciences or mathematics, the journal with the currently largest article volume came from physical sciences; the Open Journal for Quantum Science had the largest number of articles published during 2018 to 2021. All 35 journals accepted manuscripts in English. Eight of these journals also accepted manuscripts in French, of which 2 also in German, and 1 in addition to these also in Spanish and Italian. The strong presence of French as a publication language has a likely connection with Episciences, which is a French service provider to many of the overlay journals (more on this in RQ 2.10). The data concerning publisher types revealed that more than half (n=19; 54%) of the investigated overlay journals were published by groups of scientists (journals' editorial boards). In other words, the journal was published using an open repository as a manuscript source and managed through a web service (e.g. Episciences.org or Scholastica) , but no formal publisher information was found. Table 2 presents the publisher types of the investigated overlay journals. Professional, non-commercial publisher 2 (6%) University press 1 (3%) Table 2 Publisher types of the investigated overlay journals (n=35). Episciences was the most common service provider for the investigated journals. A total of 20 (57%) of the investigated journals utilised the Episciences' platform for overlay journals. Scholastica was the second most common service provider: 8 (29%) of the investigated journals utilised Scholastica's services. The majority of the investigated overlay journals (n=26; 74%) published their articles under Creative Commons (CC) licences. Table 3 presents the publishing licences used by the investigated overlay journals. It is noteworthy, that although all of the investigated journals were OA journals, 5 (14%) of the journals did not specify the licence or copyright information, which would determine the reuse terms of the published articles. Table 3 Publishing licences of the investigated overlay journals (n=35). About one third of the investigated overlay journals were indexed in DOAJ (n=11; 31%). In addition, about one third were indexed in Elsevier's Scopus database (n=12; 34%) and also in Clarivate analytics' Web of Science database (n=11; 31%). The low number of indexed journals in the prior services is at least partly explained by the fact that 17 (49%) of the journals had their first overlay volume published during 2020 or after or had not yet issued their first volume. Although a minority of journals were indexed in the Scopus database, the overall number of articles indexed in Scopus were on the rise. The analysis revealed that the Nordic journal rankings are not consistent in their classifications of the investigated overlay journals. Fourteen (40%) of the investigated journals had a Finnish Publication Forum class, and thirteen of them were considered as standard or 'basic' level journals (level 1). Respectively, 12 of the investigated journals were classified in the Norwegian Register for Scientific Journals, Series and Publisher of which 11 were ranked to level 1. Notably, only 7 of the investigated overlay journals were classified in the Danish Bibliometric Research Indicator list of journals (4 journals being classified to level 1). Appendix 1 compiles the observed journal metrics for the investigated overlay journals. The journals that had recently converted into the overlay model fared the best in the Nordic journal rankings. eLife was classified as a level 2 journal in all of the Nordic rankings and Fundamenta Informaticae was classified as level 2 in the Danish Bibliometric Research Indicator list of journals. However, as both of these journals have recently converted to the overlay model, their ranking classifications derive from their era utilising a different publishing model. From the born-overlay journals, only Logical Methods in Computer Sciences was classified as a level two journal in the Nordic journal rankings and received this classification in the Danish Bibliometric Research Indicator. The analysis revealed that overlay journals may also rank highly within the traditional journal citation metrics. Six (17%) of the investigated journals had an Impact Factor metric for the year 2020 (as released in 2021), although the Impact Factor Rankings of eLife and Fundamenta Informaticae journals were derived from their era before converting to the overlay model. The Open Journal for Quantum Science ranked the highest within their field of science specific Impact Factor rankings (see Appendix 1). This study was able to identify 35 overlay journals through an explorative data collection process. The findings suggest that overlay publishing is still in its early stage and the current amount of overlay journals is still diminutive. In comparison, the DOAJ directory currently indexes a total of 16,577 OA journals (DOAJ, 2021) and Ulrichsweb database lists a total of 24,837 open access journals. However, the findings suggest that the amount of overlay journals is on the rise. However, notable is that almost half of the investigated overlay journals (n=17, 49%) started as overlay journals from 2020 or later. This can be seen to have happened in conjunction with the growth of subject repositories and preprint uploads gaining popularity in more research disciplines. Though it has been found that many traditional journals still have unclear policies for preprint availability of submitted manuscripts (Klebel et al., 2020) , the situation has been improving over time which also gives authors increased confidence in distributing their work before or during submission to a journal. Some journals, including the portfolio of journals from the Public Library of Science (PLOS) now even incorporate direct transfer and/or facilitated posting of manuscripts to bioRxiv and medRxiv (PLOS.org, 2022) . Most of the current overlay journals represent natural sciences -computer and information sciences, physics, and mathematics in particular. In addition, notwithstanding the journals that recently transformed to the overlay model, all journals with Nordic quality rankings and Impact Factors represented the above fields. This prevalence of journals representing natural sciences is not surprising given that these fields developed the first open preprint repositories (Brown, 2001) . It seems likely that the researchers' decision to publish in overlay journals is at least partly affected by the adoption of open preprint repositories within the field. However, it is noteworthy that the HAL repository, which was the second most frequent article source for the investigated overlay journals, is a multidisciplinary repository with a scope beyond natural sciences. Furthermore, within recent years, open repositories have also been established to fields such as chemical engineering (chemRxiv), biological sciences (bioRxiv) and medical research (medRxiv) (see Chiarelli et al., 2018) . Although the most prominent overlay journals did come from natural sciences, the analysis revealed that there are emerging journals also, e.g., within the fields of basic medicine, biological sciences, and economics and business (see also Thornton & Kroeker, 2022) . Given that Tennant et al. (2017, p. 21) reported in 2017 that the overlay publishing model was not adopted beyond natural sciences, this recent development where also other fields have started to experiment with the model seems noteworthy. The citation impact comparisons between open access and subscription journals is a complex issue involving disciplinary-specific considerations (Torres-Salinas et al., 2019, p. 141 ). As pointed out by Björk and Solomon (2012) , researchers may find impactful open access journals by choosing their journals carefully. The findings of this article suggest that this principle applies also to overlay journals. The Open Journal for Quantum Science was ranked highly within the field of science specific Impact Factor rankings. Furthermore, it is noteworthy that the Impact Factor rankings are derived retrospectively (i.e. the journal's Impact Factor as released in 2020 is based on the citation data of 2017-2018). As 15 (43%) of the investigated journals were founded after 2017, this excludes them from even plausibly having an Impact Factor metric released in 2020. The detailed analysis of the costs of individual overlay articles or the model are beyond the scope of this article. The operating costs of arXiv (Cornell University, 2022) and overlay journal web services, such as Scholastica (2021) and Episciences.org (2021) , suggest that the overlay model is cost efficient when compared to the traditional subscription-based publishing (Schimmer et al., 2015; Grossmann & Brembs, 2021) . However, given the current diminutive amount of overlay journals, the larger scale costsaving effects of the model seem to be linked to the willingness of researchers to adopt this model of publishing in the future. It is noteworthy that nearly all of the investigated journals did not charge fees from the authors or readers, which is at least partly explained by the cost-effectiveness of the overlay publishing model. In a recent study of preprints available in bioRxiv, Anderson (2020) found that the majority of manuscripts had been uploaded to the repository close to or after submission to a journal, suggesting that pre-prints were not dominantly used as a mechanism for pre-submission feedback. Based on the study a third of preprints had not been published as peer-reviewed journal articles within two years of upload. Considering these results in the context of the process flows that could be observed from the policies of the overlay journals included in this study there is not anything inherently in the common implementation of overlay model publishing that would change this pattern. Submission to a repository can be done just before submitting to the journal (most journals just ask for a link to the preprint in the submission form) or does not need to happen at all if the journal also accepts direct submissions but uses repository functions for archiving published content (as was the case for 9 out of 35 journals). This combined with the low presence of open peer review practices (2 out of 35 journals with some transparency of the review process) among these journals suggest that the model does not, as it is commonly implemented today, provide a radical change to the open science practices as part of the pre-publication and review process. The overlay journal movement is still in its early stages, and it could be argued that it is less of a cohesive movement and rather better to be considered a diverse set of ways through which journals integrate repository archival into their workflows -sometimes to facilitate open science practices early on in the submission stages, sometimes just for robust archival of accepted and published content. The current presence of overlay journals is diminutive compared to the overall number of open access journals, but the number is currently increasing. The research fields for overlay journals are skewed towards journals within computer science and informatics, mathematics, and the physical sciences. In addition, recently established overlay journals also include medical, biological and multidisciplinary research. Overlay journals are commonly published by groups of scientists rather than formal organisations and overlay journals may also rank highly within the traditional journal citation metrics. Nearly all of the investigated overlay journals did not charge fees from the authors or readers, which is at least partly explained by the cost-effectiveness of the overlay publishing model. Both the adoption of open access preprint repositories and researchers' willingness to publish in overlay journals will determine the model's wider impact on scientific publishing. 14 (40%) 7 (20%) 12 (34%) 8 (29%)* 6 (17%)* 6 (17%)* Biological sciences (n=1); Computer and information sciences (n=5); Economics and business (n=1); Mathematics (n=6); Physical sciences (n=3) bioRxiv: Trends and analysis of five years of preprints Open access versus subscription journals: a comparison of scientific impact The E-volution of preprints in the scholarly communication of physicists and astronomers Preprints and scholarly communication: An exploratory qualitative study of adoption, practices, drivers and barriers Web of Science Core Collection Directory of Open Access Journals Scopus database Celebrating 20 Years of Open Access and Innovation at JMIR Publications Publication forum The evolving role of preprints in the dissemination of COVID-19 research and their impact on the science communication landscape Current market rates for scholarly publishing services The changed-and changinglandscape of serials publishing: Review of the literature on emerging models Peer review and preprint policies are unclear at most major journals Delayed open access: An overlooked high-impact category of openly available scientific literature Open is not forever: a study of vanished open access journals arXiv E-Prints and the Journal of Record The Bibliometric Research Indicator Norwegian Directorate for Higher Education and Skills Open Journals. A collection of open source, open access journals Revised field of science classification for the Frascati manual Journals and repositories: an evolving relationship? Learned publishing Disrupting the subscription journals' business model for the necessary large-scale transformation to open access Why Scholastica The deconstructed journal -a new model for academic publishing The journal as an overlay on preprint databases A multi-disciplinary perspective on emergent and future innovations in peer review Overlay Journals: Overlooked or Emergent Disentangling gold open access Springer handbook of science and technology indicators Ulrichsweb: global serials directory The authors thank Dr. Tommi Tenkanen for his important comments and feedback considering this work. The authors also thank the Scholastica and PubPub teams for kindly providing information of their overlay journals. The collected data are provided as open data at https://doi.org/10.5281/zenodo.6420517 to facilitate future research. Appendix 1: Nordic journal rankings, Journal Citation impact and Impact Factor metrics of the investigated overlay journals (n=35). *= The journals of eLife, Fundamenta Informaticae, Journal of Groups, Complexity, Cryptology, and Journal of Philosophical Economics were recently converted to the overlay model, and their Journal Citation Impact, Impact Factors and Impact Factor Rankings are derived from their era before converting to the overlay model.