Learned Publishing最新文献

{"title":"Clinical Evidence Behind a Paywall: An Analysis of Randomised Clinical Trials Included in Cochrane Reviews","authors":"Leonardo J. Uribe-Cavero,&nbsp;Ana Brañez-Condorena,&nbsp;Alexander M. Parra-Huaroto,&nbsp;Patricia J. Vera-Maccha,&nbsp;Alvaro Taype-Rondan","doi":"10.1002/leap.2058","DOIUrl":"10.1002/leap.2058","url":null,"abstract":"<p>Open access to health evidence is essential for informed decision-making, yet many randomised controlled trials (RCTs) remain behind paywalls. This study assessed the accessibility of RCTs cited in Cochrane intervention systematic reviews, focusing on open access availability and alternative access options. We analysed 491 RCTs from a random sample of 50 Cochrane reviews published in early 2023. Overall, 43.0% required payment for access, but this proportion declined over time, from 60.0% for articles published between 1980 and 1999 to 20.7% for those from 2020 to 2023. Among paywalled articles, the lowest fee provided 24-h access in 37.0% of cases. However, 49.8% of these studies were available in full text via Google Scholar, and 95.3% of the remaining could be accessed through Sci-Hub. Despite the persistent presence of paywalls, access barriers have decreased over time, and alternative platforms provide widespread availability of restricted articles.</p>","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-04-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2058","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147686841","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Scoping Reviews Should Describe—Not Score","authors":"Tove Godskesen","doi":"10.1002/leap.2057","DOIUrl":"10.1002/leap.2057","url":null,"abstract":"<p>Recent calls for increased transparency in scoping reviews, such as those by Ang et al. (<span>2026</span>) in <i>Learned Publishing</i>, have led to suggestions to use tools like the Quality Assessment with Diverse Studies (QuADS) to enhance reporting clarity.</p><p>Originally developed to appraise reporting in mixed-methods research (Harrison et al. <span>2021</span>), QuADS was not designed for the broader, more inclusive goals of scoping reviews. While well-intentioned, arguments for applying QuADS to scoping reviews risk conceptual misalignment and methodological overreach. Instead of enhancing clarity, it may inadvertently distort the purpose and strengths of the method.</p><p>Efforts to improve reporting and transparency in scoping reviews are vital. But the routine use of evaluative tools like QuADS assumes a narrow definition of quality that conflicts with the exploratory and epistemically inclusive aims of scoping reviews.</p><p>Instead of applying rigid methodological frameworks, we should embrace the diverse evidence base of scoping reviews, ensuring transparency through reflexive and context-sensitive reporting. Tools should enable, not constrain, the method.</p><p>The sole author contributed to all aspects of this manuscript.</p><p>The author has nothing to report.</p><p>The author has nothing to report.</p><p>The author has nothing to report.</p><p>The author declares no conflicts of interest.</p><p>Data sharing not applicable to this article as no datasets were generated or analysed during the current study.</p>","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2057","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147686790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From Findings to Meaning: A Strategic Framework for the Discussion Section","authors":"Farrokh Habibzadeh","doi":"10.1002/leap.2050","DOIUrl":"10.1002/leap.2050","url":null,"abstract":"&lt;p&gt;The Discussion section is often the most intellectually demanding component of the IMRaD (Introduction, Methods, Results and Discussion) structure (Cals and Kotz &lt;span&gt;2013&lt;/span&gt;; Cohen et al. &lt;span&gt;2016&lt;/span&gt;). While the preceding sections follow relatively conventional formats, they serve distinct and deliberately constrained roles. The Introduction establishes the current state of knowledge and culminates in the specific research question or hypothesis. The Methods section details the experimental materials and procedures, and the Results section presents the findings objectively. Crucially, these sections are reserved for reporting established knowledge, methodological choices and empirical data—not for interpretation (Cals and Kotz &lt;span&gt;2013&lt;/span&gt;). The Discussion, therefore, serves as the primary venue for interpretation and synthesis. It is here that authors transform data into understanding, elucidating the meaning and implications of their findings—where numbers become meaning (Cohen et al. &lt;span&gt;2016&lt;/span&gt;).&lt;/p&gt;&lt;p&gt;A well-crafted Discussion begins with a focused interpretation of the specific results and broadens to integrate them into the existing scientific landscape, comparing and contrasting with relevant literature (Kearney &lt;span&gt;2017&lt;/span&gt;). Whereas the Introduction follows an inverted pyramid structure—moving from general context to a specific research question—the Discussion proceeds in the opposite direction, from specific findings toward their general implications, resembling an upright pyramid (Bavdekar &lt;span&gt;2015&lt;/span&gt;). This section acts as the crucible for generating new, evidence-supported ideas. The absence of a universal template, despite general guidelines, is what makes crafting a compelling Discussion particularly challenging, yet it affords authors the freedom to express their insights in the most appropriate way (Conn &lt;span&gt;2017&lt;/span&gt;).&lt;/p&gt;&lt;p&gt;The function of the Discussion extends beyond summarising results. It interprets the obtained data, compares them with previous findings, outlines the implications of the new knowledge, identifies the strengths and limitations of the study and proposes future research directions (Cals and Kotz &lt;span&gt;2013&lt;/span&gt;; Höfler et al. &lt;span&gt;2018&lt;/span&gt;). Nevertheless, many manuscripts falter here; authors often reiterate results, speculate without evidence, or list literature mechanically, merely comparing their findings with those of others without explaining the probable reasons for similarities or differences (Cohen et al. &lt;span&gt;2016&lt;/span&gt;; Kearney &lt;span&gt;2017&lt;/span&gt;).&lt;/p&gt;&lt;p&gt;Herein, I provide a conceptual and structural framework to help authors craft lucid, persuasive Discussions while avoiding common pitfalls (Conn &lt;span&gt;2017&lt;/span&gt;).&lt;/p&gt;&lt;p&gt;The Discussion has three overarching purposes: (1) interpretation—explaining what the findings mean; (2) integration—placing them in the context of prior knowledge; and (3) implication—showing why they matter and what should follow (C","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2050","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147579860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Scholarly Communications in 2025: An Aerial Evaluation of a System Challenged by AI and Much More","authors":"David Nicholas,&nbsp;Abdullah Abrizah,&nbsp;Eti Herman,&nbsp;Jorge Revez,&nbsp;John Akeroyd,&nbsp;Blanca Rodríguez-Bravo,&nbsp;Marzena Swigon,&nbsp;David Clark,&nbsp;Tatyana Polezhaeva","doi":"10.1002/leap.2056","DOIUrl":"10.1002/leap.2056","url":null,"abstract":"<p>Using data obtained from the 2025 round of the Harbingers project on early career researchers (ECRs), artificial intelligence (AI) and scholarly communications, we provide an overarching (aerial) analysis of the AI-impacted scholarly communications system. It covers nearly 20 communication aspects, including metrics, peer review, which saw 62 ECRs interviewed from 6 countries and from a range of subjects. To produce such a wide panorama in a single paper the focus is on the interview questions that provided summarisation, codification and quantification of the data. The data was analysed by country, age, gender and subject, where relevant and significant. Quotes are also used to support the quantitative data. There is widespread agreement that citations remain the main reputational currency, that traditional publishing outlets are here to stay, that peer review badly needs fixing and, most importantly, that AI is impacting on a very wide front across the scholarly enterprise. Overhanging it all is the worry that the quality of research output is being undermined, through bad actors and AI ghostwriters. Findings are based on a relatively small, convenience sample, so they should not be regarded as definitive, rather as pointers. Furthermore, an extensive literature review shows that this is a rare study of ECRs.</p>","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2056","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147579890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing, Understanding and Adoption of the Contributor Roles Taxonomy (CRediT)","authors":"Mohammad Hosseini,&nbsp;Simon Kerridge,&nbsp;Liz Allen,&nbsp;Veronique Kiermer,&nbsp;Kristi Holmes","doi":"10.1002/leap.2048","DOIUrl":"https://doi.org/10.1002/leap.2048","url":null,"abstract":"&lt;p&gt;Introduced for widespread use in 2015, the Contributor Roles Taxonomy (CRediT) offers 14 standard roles to classify scholarly contributions to research outputs (Allen et al. &lt;span&gt;2014&lt;/span&gt;). In 2022, CRediT was formalized as a standard by the American National Standards Institute (ANSI) and the National Information Standards Organisation (NISO), facilitating a sustained support system.&lt;/p&gt;&lt;p&gt;Thus far, several key academic workflows, including journal submission and review systems (e.g., Editorial Manager, PubSweet, ScholarOne, ReView, and the Open Journals System) have incorporated CRediT, resulting in its adoption by hundreds of journals across many major publishers including PLOS, SAGE, Springer Nature, Wiley, Frontiers, and Elsevier. That said, a recent retrospective study showed that CRediT has been inconsistently integrated among publishers, research domains and countries (Allen et al. &lt;span&gt;2025a&lt;/span&gt;). For example, while PLOS has fully implemented CRediT and mandates it across its portfolio, other publishers started much smaller and without a mandate. Elsevier, MDPI, Springer Nature, and Frontiers are the four publishers with the highest number of publications that include CRediT roles. In terms of research domains, publications in Engineering, Biomedical and Clinical Sciences, Biological Science and Chemical Sciences have the highest rate of CRediT inclusion. When considering the country of the corresponding author, the largest number of articles with CRediT information comes from China, followed by those based in the U.S., India, and the U.K. According to this study, in 2024, 22.5% (&lt;i&gt;n&lt;/i&gt; = 848,841) of original research articles, preprints, and conference papers with available full text indexed on the Dimensions platform employed the CRediT taxonomy to specify author contributions.&lt;/p&gt;&lt;p&gt;Further integration of CRediT in the scholarly landscape faces both technical and social challenges. Examples of technical challenges include difficulties of incorporating the taxonomy in legacy submission and hosting platforms that were not originally designed to support structured contributor role metadata. Moreover, the lack of standardised implementation of CRediT across publishers and platforms leads to inconsistent metadata. For example, while PLOS has integrated CRediT into its submission workflow, many other publishers and journals collect CRediT information only as narrative declarations at the end of a manuscript, meaning that these contributions are not carried forward as structured metadata. There is also limited interoperability between manuscript submission systems, publishing platforms, and downstream metadata services (e.g., indexing and discovery systems), which makes tallying contributions complicated. Social challenges include training researchers on how to use CRediT and for what purpose (i.e., “CRediT is NOT intended to define what constitutes authorship – but instead to describe the specific contributions of authors and o","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2048","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147562348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing, Understanding and Adoption of the Contributor Roles Taxonomy (CRediT)","authors":"Mohammad Hosseini,&nbsp;Simon Kerridge,&nbsp;Liz Allen,&nbsp;Veronique Kiermer,&nbsp;Kristi Holmes","doi":"10.1002/leap.2048","DOIUrl":"https://doi.org/10.1002/leap.2048","url":null,"abstract":"&lt;p&gt;Introduced for widespread use in 2015, the Contributor Roles Taxonomy (CRediT) offers 14 standard roles to classify scholarly contributions to research outputs (Allen et al. &lt;span&gt;2014&lt;/span&gt;). In 2022, CRediT was formalized as a standard by the American National Standards Institute (ANSI) and the National Information Standards Organisation (NISO), facilitating a sustained support system.&lt;/p&gt;&lt;p&gt;Thus far, several key academic workflows, including journal submission and review systems (e.g., Editorial Manager, PubSweet, ScholarOne, ReView, and the Open Journals System) have incorporated CRediT, resulting in its adoption by hundreds of journals across many major publishers including PLOS, SAGE, Springer Nature, Wiley, Frontiers, and Elsevier. That said, a recent retrospective study showed that CRediT has been inconsistently integrated among publishers, research domains and countries (Allen et al. &lt;span&gt;2025a&lt;/span&gt;). For example, while PLOS has fully implemented CRediT and mandates it across its portfolio, other publishers started much smaller and without a mandate. Elsevier, MDPI, Springer Nature, and Frontiers are the four publishers with the highest number of publications that include CRediT roles. In terms of research domains, publications in Engineering, Biomedical and Clinical Sciences, Biological Science and Chemical Sciences have the highest rate of CRediT inclusion. When considering the country of the corresponding author, the largest number of articles with CRediT information comes from China, followed by those based in the U.S., India, and the U.K. According to this study, in 2024, 22.5% (&lt;i&gt;n&lt;/i&gt; = 848,841) of original research articles, preprints, and conference papers with available full text indexed on the Dimensions platform employed the CRediT taxonomy to specify author contributions.&lt;/p&gt;&lt;p&gt;Further integration of CRediT in the scholarly landscape faces both technical and social challenges. Examples of technical challenges include difficulties of incorporating the taxonomy in legacy submission and hosting platforms that were not originally designed to support structured contributor role metadata. Moreover, the lack of standardised implementation of CRediT across publishers and platforms leads to inconsistent metadata. For example, while PLOS has integrated CRediT into its submission workflow, many other publishers and journals collect CRediT information only as narrative declarations at the end of a manuscript, meaning that these contributions are not carried forward as structured metadata. There is also limited interoperability between manuscript submission systems, publishing platforms, and downstream metadata services (e.g., indexing and discovery systems), which makes tallying contributions complicated. Social challenges include training researchers on how to use CRediT and for what purpose (i.e., “CRediT is NOT intended to define what constitutes authorship – but instead to describe the specific contributions of authors and o","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2048","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147567239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sharpening the Pencil, Not Replacing the Hand: De-Stigmatising AI Use in Research Writing","authors":"Aaron Opdyke","doi":"10.1002/leap.2055","DOIUrl":"https://doi.org/10.1002/leap.2055","url":null,"abstract":"&lt;p&gt;Artificial Intelligence (AI) tools are becoming increasingly embedded in research writing workflows. From language refinement to structural reorganisation, generative AI offers considerable aid to scholars engaged in the labour of writing. Yet as these tools proliferate, so too do questions about ethics, authorship, and integrity (Butson and Spronken-Smith &lt;span&gt;2024&lt;/span&gt;). At the heart of these questions lies a tension: while AI promises clarity and efficiency, it also elicits discomfort about ownership, transparency, and what counts as legitimate scholarly work.&lt;/p&gt;&lt;p&gt;A recent study by Kwon (&lt;span&gt;2025a&lt;/span&gt;) found that 90% of researchers now consider it acceptable to use generative AI to edit their writing, yet many remain unsure how, and even whether, to disclose that use. This gap between acceptance and transparency underscores the unease that surrounds AI in research writing: the fear that acknowledging AI assistance might be interpreted as cutting corners or relinquishing intellectual control. This is especially evident in university settings, where institutional policies often create unequal permissions for AI use, for example, enthusiastically embracing it (McDonald et al. &lt;span&gt;2025&lt;/span&gt;), but restricting its use. These divergent expectations can surface deeper assumptions about scientific labour, expertise, and trust, and contribute to confusion, suspicion, and inconsistency in how AI is perceived across the academic community.&lt;/p&gt;&lt;p&gt;Rather than contribute to the growing list of prohibitions or fears—each of which brings its own ethical, methodological, and disciplinary complexities—this piece advances a more provocative proposition that seems to be implicitly acknowledged, yet remains less visible in current discourse: that far from undermining research integrity, using AI, when done responsibly, can actually enhance it. Framed as a form of reflexive autoethnography (Adams et al. &lt;span&gt;2021&lt;/span&gt;), this perspective draws on a series of recent interactions with students, colleagues, and collaborators, which revealed a recurring pattern of unease, hesitation, and misinterpretation surrounding AI-assisted writing. These conversations often centred on questions of authorship, ownership, and fairness, particularly around assumptions that AI use is a shortcut or that it diminishes the human contribution. Such reactions reflect not just a gap in policy guidance, but a wider cultural tension marked as much by silence as by critique—a tension that this piece seeks to unpack.&lt;/p&gt;&lt;p&gt;The unease surrounding generative AI in academic writing is often less about the tool itself and more about what its use is assumed to represent. In my own interactions, there have been markedly different attitudes towards different tools: ChatGPT, for instance, tends to evoke suspicion or carry negative connotations, whereas similar AI tools are met with ambivalence or pass without comment. This disparity suggests that perceptions are shaped less by functi","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2055","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sharpening the Pencil, Not Replacing the Hand: De-Stigmatising AI Use in Research Writing","authors":"Aaron Opdyke","doi":"10.1002/leap.2055","DOIUrl":"https://doi.org/10.1002/leap.2055","url":null,"abstract":"&lt;p&gt;Artificial Intelligence (AI) tools are becoming increasingly embedded in research writing workflows. From language refinement to structural reorganisation, generative AI offers considerable aid to scholars engaged in the labour of writing. Yet as these tools proliferate, so too do questions about ethics, authorship, and integrity (Butson and Spronken-Smith &lt;span&gt;2024&lt;/span&gt;). At the heart of these questions lies a tension: while AI promises clarity and efficiency, it also elicits discomfort about ownership, transparency, and what counts as legitimate scholarly work.&lt;/p&gt;&lt;p&gt;A recent study by Kwon (&lt;span&gt;2025a&lt;/span&gt;) found that 90% of researchers now consider it acceptable to use generative AI to edit their writing, yet many remain unsure how, and even whether, to disclose that use. This gap between acceptance and transparency underscores the unease that surrounds AI in research writing: the fear that acknowledging AI assistance might be interpreted as cutting corners or relinquishing intellectual control. This is especially evident in university settings, where institutional policies often create unequal permissions for AI use, for example, enthusiastically embracing it (McDonald et al. &lt;span&gt;2025&lt;/span&gt;), but restricting its use. These divergent expectations can surface deeper assumptions about scientific labour, expertise, and trust, and contribute to confusion, suspicion, and inconsistency in how AI is perceived across the academic community.&lt;/p&gt;&lt;p&gt;Rather than contribute to the growing list of prohibitions or fears—each of which brings its own ethical, methodological, and disciplinary complexities—this piece advances a more provocative proposition that seems to be implicitly acknowledged, yet remains less visible in current discourse: that far from undermining research integrity, using AI, when done responsibly, can actually enhance it. Framed as a form of reflexive autoethnography (Adams et al. &lt;span&gt;2021&lt;/span&gt;), this perspective draws on a series of recent interactions with students, colleagues, and collaborators, which revealed a recurring pattern of unease, hesitation, and misinterpretation surrounding AI-assisted writing. These conversations often centred on questions of authorship, ownership, and fairness, particularly around assumptions that AI use is a shortcut or that it diminishes the human contribution. Such reactions reflect not just a gap in policy guidance, but a wider cultural tension marked as much by silence as by critique—a tension that this piece seeks to unpack.&lt;/p&gt;&lt;p&gt;The unease surrounding generative AI in academic writing is often less about the tool itself and more about what its use is assumed to represent. In my own interactions, there have been markedly different attitudes towards different tools: ChatGPT, for instance, tends to evoke suspicion or carry negative connotations, whereas similar AI tools are met with ambivalence or pass without comment. This disparity suggests that perceptions are shaped less by functi","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2055","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI And the Editors' Ghost: Who Is the Writer Now?","authors":"David Clark,&nbsp;David Nicholas,&nbsp;Abdullah Abrizah,&nbsp;John Akeroyd,&nbsp;Jorge Revez,&nbsp;Blanca Rodríguez-Bravo,&nbsp;Marzena Swigon,&nbsp;Tatyana Polezhaeva,&nbsp;Anne Gere,&nbsp;Eti Herman","doi":"10.1002/leap.2051","DOIUrl":"https://doi.org/10.1002/leap.2051","url":null,"abstract":"<p>This an exploration of the use of AI in research and writing. It builds upon the ‘Harbingers’ project, an international and longitudinal study of early career researchers (ECRs) and scholarly communication. In the fourth phase of the project, we returned to the theme of AI, in particular AI as ‘ghostwriter’. Our sources are transcripts of conversational, open-form interviews with over 60 ECRs from Britain, Malaysia, Poland, Portugal, Spain, Russia, and other countries. For an initial analysis of the transcripts, we used Google NotebookLM. An overarching and thematic summary of the data was produced in minutes, that would otherwise have occupied our research team for weeks. The unprompted text, immediately plausible and coherent, was regarded by all national interviewers as impressive. Here, using a relatively small, convenience sample, we compare the AI generated summaries both against our original data and those first impressions. We reflect upon our own experience of using AI and that of our interviewees. This paper is about how we used AI as an experiment, our reaction to it, how that chimes, resonates, echoes the experiences of the ECRs. It is a calibration for our future data analysis.</p>","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2051","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147565982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards a DOI-First Referencing Model: Opportunities, Limitations and Implications for Scholarly Publishing","authors":"Mazhar Mushtaq","doi":"10.1002/leap.2052","DOIUrl":"https://doi.org/10.1002/leap.2052","url":null,"abstract":"&lt;p&gt;&lt;i&gt;Why Are We Still Formatting References Manually?&lt;/i&gt; Despite transformative advances in artificial intelligence (AI), one aspect of academic publishing remains strikingly outdated: reference formatting. Most journals still require authors to use discipline-specific or journal-specific citation styles, such as Vancouver, APA, AMA, MLA or proprietary variations, none of which align with their original styles. These formats differ in punctuation, abbreviation rules, capitalisation patterns and ordering conventions. Although these stylistic choices add no scientific value, they impose a persistent and unnecessary burden on authors, reviewers, editors and publishers.&lt;/p&gt;&lt;p&gt;&lt;i&gt;The Digital Paradox: AI-Assisted Writing versus Manual Referencing&lt;/i&gt;: AI-assisted writing tools now perform tasks once considered impossible: automated literature searches, structure-aware drafting, screening of systematic-review records, summarisation of complex findings and even preliminary manuscript generation. Yet researchers still spend hours adjusting italics, checking superscripts, reordering author lists and correcting journal abbreviations, activities that neither improve quality nor enhance reproducibility. This contrast highlights an inconsistency between our modern digital tools and the analogue habits preserved within editorial workflows. These practices persist even though they contribute nothing to scientific quality, reproducibility or clarity. Instead, they reflect a legacy workflow inherited from print-era publishing. Furthermore, the journal's publishers have adopted new tools for manuscript submission. However, reference lists remain dependent on manual formatting or on citation software that often introduces inconsistencies and requires substantial author intervention.&lt;/p&gt;&lt;p&gt;The DOI (Digital Object Identifier) already provides a universal, machine-readable, permanent and unambiguous identifier for scholarly objects, operating as the research world's equivalent of the QR code: compact, unique and instantly retrievable. A DOI retrieves full citation metadata, including authors, title, journal, year, pages, publisher and links to the version of record. Databases, citation APIs, indexing services and AI systems all rely on DOI-based metadata rather than human-formatted references. These data can be retrieved instantly through registration agencies such as CrossRef and DataCite (Lee and Stvilia &lt;span&gt;2012&lt;/span&gt;; Starr and Gastl &lt;span&gt;2011&lt;/span&gt;; Wiley &lt;span&gt;2014&lt;/span&gt;).&lt;/p&gt;&lt;p&gt;Therefore, I propose that journals adopt a &lt;i&gt;universal DOI-first citation format&lt;/i&gt;, whereby each cited source is listed simply in one line as:&lt;/p&gt;&lt;p&gt;DOI: 10.xxxx/xxxxx &lt;i&gt;OR&lt;/i&gt; https://doi.org/10.xxxx/xxxxx&lt;/p&gt;&lt;p&gt;Without the use of any italics, abbreviations or stylistic rules, like the use of three authors, six authors or et al. Similarly, the use of superscripts, square brackets, bold initials, italics, it's, all mixed with commas, semicolons, years before or after, names in","PeriodicalId":51636,"journal":{"name":"Learned Publishing","volume":"39 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/leap.2052","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147566056","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}