How many academic documents are visible and freely available on the Web?
|
|
Scooped by
Gilbert C FAURE
onto Notebook or My Personal Learning Network November 29, 2015 1:54 PM
|
|
|
Scooped by
Gilbert C FAURE
onto Notebook or My Personal Learning Network November 29, 2015 1:54 PM
|
How many academic documents are visible and freely available on the Web?
Your new post is loading...
Your new post is loading...
|
|
|
Scooped by
Gilbert C FAURE
Today, 8:49 AM
|
Six layers of research methodology
Research Philosophy: Why and how knowledge is understood.
Research Approach: How theory and evidence are connected.
Research Strategy: What overall research design is adopted.
Methodological Choice: What type of research is conducted.
Time Horizon: When and over what period data is collected.
Techniques and Procedures: This is the innermost layer and explains the practical methods used for collecting and analysing data.
|
|
Scooped by
Gilbert C FAURE
November 20, 9:59 AM
|
When Academic Citations Become a Controversy: The Daniel Balsalobre Case
Daniel Balsalobre, a Spanish environmental economics researcher affiliated with the University of Castilla-La Mancha, has become the subject of a major academic publishing controversy.
According to an investigative report published by EL PAÍS on November 17, 2026, six of his papers have been retracted over concerns involving citation manipulation, peer-review irregularities, and undisclosed conflicts of interest.
The details are striking.
Balsalobre reportedly has more than 230 publications and 25,000 citations. His publication rate increased from around two papers per year in 2015 to approximately one paper every five working days in 2023.
One particularly controversial allegation concerns his peer-review activities. He reportedly accepted more than 650 review invitations. In one case, while reviewing a paper about ChatGPT in higher education, he allegedly asked two Indian researchers to include seven of his own publications on environmental economics in their references.
The investigation also identified connections with several international researchers whose papers had faced retractions.
Even more surprisingly, 170 papers citing his research have themselves reportedly been retracted.
Balsalobre denies deliberate misconduct, although he acknowledges suggesting citations to his own work in some reviews. His university has initiated an investigation.
One important question:
If citation counts can be influenced by academic networks, peer-review relationships, and requests for self-citation, how reliable are they as a measure of the world's most influential researchers?
#AcademicPublishing #ResearchIntegrity #PeerReview #CitationMetrics #HigherEducation #Scientometrics | 26 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 20, 5:20 AM
|
The researchers estimated that if treatment for all eligible individuals were fully aligned with U.S. or E.U. guidelines, including the use of non-statin LDL-lowering drugs in many cases, median levels of LDL cholesterol would drop sharply, reducing the risk of major cardiovascular events in the U.S. by up to 27%.
“Several factors account for the gaps that we document,” says researcher Alexander. “They include differences in clinician training, patient preferences, barriers to accessing care, financial incentives that don’t always support best practices, and the difficulty of putting clinical guidelines into practice in busy, real-world settings.”
Bringing actual treatment closer to what guidelines recommend could be achieved through various measures including better patient education on the benefits of treatment for those who know they have high LDL-cholesterol levels, and better screening for everyone else, the researchers say.
|
|
Scooped by
Gilbert C FAURE
November 19, 4:00 AM
|
𝗠𝗼𝘀𝘁 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗽𝗮𝗽𝗲𝗿𝘀 𝗮𝗿𝗲𝗻’𝘁 𝗵𝗮𝗿𝗱 𝘁𝗼 𝗿𝗲𝗮𝗱 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝗼𝗳 𝘁𝗵𝗲 𝘀𝗰𝗶𝗲𝗻𝗰𝗲 𝘁𝗵𝗲𝘆’𝗿𝗲 𝗵𝗮𝗿𝗱 𝘁𝗼 𝗿𝗲𝗮𝗱 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝗼𝗳 𝘁𝗵𝗲 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲
A research paper has an anatomy. When each section knows its job, the paper flows from the first line to the last. When sections blur together, readers and reviewers lose the thread. The Fish Model gives you a simple picture to hold in your head: one body, seven parts, each doing exactly one thing.
Here’s the 𝗙𝗶𝘀𝗵 𝗠𝗼𝗱𝗲𝗹 𝗼𝗳 𝗮 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗽𝗮𝗽𝗲𝗿 👇
🐟 𝟭. 𝗔𝗯𝘀𝘁𝗿𝗮𝗰𝘁 — The head. State the problem, the methodology used, a summary of findings, and the conclusion.
🐟 𝟮. 𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 — Set the scene with the literature review. What is the problem or burden? What is known and what is unknown? What is your aim and hypothesis?
🐟 𝟯. 𝗠𝗲𝘁𝗵𝗼𝗱𝗼𝗹𝗼𝗴𝘆 — Explain the study design, study population, and data source. Define your risk factors (exposures) and outcomes, and describe your statistical analysis.
🐟 𝟰. 𝗥𝗲𝘀𝘂𝗹𝘁𝘀 — What did you find out? Present your primary and secondary findings, with data represented in tables and figures.
🐟 𝟱. 𝗗𝗶𝘀𝗰𝘂𝘀𝘀𝗶𝗼𝗻 — Interpret and contextualize your findings in relation to existing knowledge. Cover the significance, implications, and limitations.
🐟 𝟲. 𝗖𝗼𝗻𝗰𝗹𝘂𝘀𝗶𝗼𝗻 — The take-home message. Keep it short and memorable.
🐟 𝟳. 𝗥𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲𝘀 — The existing knowledge you built this work upon.
Notice the shape: the paper starts broad at the head, narrows through the methods and results at the body, then widens again in the discussion before closing with a clear tail. That’s why the model works, it shows you where your paper should expand and where it should stay tight. 🎓
A strong paper isn’t written section by section in isolation, it’s built as one body where every part supports the next. ✍️
Save this as your paper-writing blueprint before your next draft. 📌
Want more handbooks and guides like this? Visit Research Crave Academy 👇
🌐 Researchcraveacademy.com
Need help with your research paper, structure, or journal submission? Reach out 👇
📧 asma@researchcrave.com
🌐 www.researchcrave.com
whatsapp: https://wa.link/bbvf22
#ResearchPaper #AnatomyOfAPaper #FishModel #AcademicWriting #ResearchWriting #AcademicWritingTips #PaperStructure #HowToWriteAResearchPaper #AcademicPublishing #JournalPublication #ResearchMethodology #AcademicResearch #PhDLife #PhDStudent #PhDJourney #PhDSupport #PhDCommunity #GradSchool #HigherEducation #DoctoralStudies #PostgraduateLife #ResearchSkills #ScholarlyWriting #ResearchCrave #ResearchCraveAcademy| 11 commentaires sur LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 19, 3:58 AM
|
What if the collapse of the journal article as academic currency is actually a good thing?
I've seen a lot of concerns lately about academics using AI to write scientific papers and peer reviewers using AI to review those same papers. AI writing papers. AI reviewing papers. Humans collecting academic credit.
Anna Wexler's recent article in The Atlantic, "The Currency of the Journal Article Is Collapsing," raises legitimate concerns about what AI is doing to scientific publishing.
https://lnkd.in/eXJH7gG9
Here's another take:
Perhaps AI isn't destroying the value of academic publishing. Maybe it's exposing how distorted our incentives have become.
Scientific journals were developed to communicate discoveries, exchange ideas, and advance knowledge. Somewhere along the way, publications became the currency of academia. Publish or perish. Count the papers. Chase the impact factors. Secure the next promotion.
Now, AI is making it easier to produce manuscripts. If generating a manuscript becomes increasingly effortless, why are we still using the number of manuscripts someone produces as a measure of their academic contribution?
Meanwhile, a multibillion-dollar publishing industry profits from this system. Universities and taxpayers fund much of the research. Academics write and review the manuscripts, often on their own time. Then institutions pay publishers for access or to make that research openly available. Quite the business model.
To be clear, I'm not suggesting we abandon scientific rigor or the need to critically evaluate and disseminate research. AI-generated nonsense masquerading as scholarship is a genuine problem.
But the value of scientific work should lie in the knowledge it produces, not the number of manuscripts it generates or the prestige of the journal that publishes them.
I'd be perfectly content to see publications lose their status as the dominant currency of academic achievement.
That does bring up another question.
If we stopped counting publications as the primary measure of academic success, what would you want academia to reward instead?
Ken Milne Andrew Tagg Tessa Davis Damian Roland Ioannis Koutroulis| 11 commentaires sur LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 19, 3:55 AM
|
𝐀𝐜𝐚𝐝𝐞𝐦𝐢𝐜 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐏𝐚𝐩𝐞𝐫 𝐒𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞
A flowchart illustrating the five key sections of a research paper: Abstract, Introduction, Methods, Results, and Discussion, with their essential components.
𝐃𝐌 𝐮𝐬 𝐢𝐟 𝐲𝐨𝐮 𝐧𝐞𝐞𝐝 𝐚𝐧𝐲 𝐫𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐚𝐬𝐬𝐢𝐬𝐭𝐚𝐧𝐜𝐞!
#ResearchPaper #AcademicWriting #PhDResearch #ResearchMethodology #ResearchScholars #ThesisWriting #AcademicResearch | 10 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 19, 3:48 AM
|
Bias on Wikipedia isn’t the result of who the editors are. It’s about what they do.
That’s my biggest takeaway from wikihistories, our three-year ARC-funded project on how Wikipedia represents Australian people, places and historical events. Last week our team presented the findings, and what they mean for Wikimedians, at a seminar hosted by the great team at Wikimedia Australia.
It’s tempting to explain Wikipedia’s biases by who its editors are. About 80% of them are male, for example. But we found that everyday practices matter more, especially rules and policies and how they’re applied. Those practices shape what gets represented and what stays hidden or obscured.
Take place name changes in Australia. Discussions became so hostile that experienced editors avoided them, which often left the debates to the most extreme voices. Rules about which sources count as “reliable” shaped the outcomes too.
But Wikipedians don’t only apply rules. They can also make them. Australian editors could set local guidelines for place names. These could cover how long to wait after an official renaming before changing an article’s title, or how much weight to give a government announcement made after public consultation compared with how widely the media has adopted the new name. Rules like these might have shortened the long wait to rename K’gari, and made the process less chaotic and uncertain.
This local rule-making is where I now see Wikipedia’s greatest opportunity. Editors can work not only to add the people and things we care about, but also to change how we produce knowledge together.
There are strong pressures to centralise Wikipedia’s rules and practices. Meanwhile, communities outside the US, outside English Wikipedia and away from the most-visited articles are developing their own norms and cultures. Which force wins is far from settled.
I’m more convinced than ever that Wikipedia is not one thing. Its cultures, histories and effects need to be assessed locally, where “local” means more than geography.
Thanks to Wikimedia Australia for hosting us and Elliott Bledsoe for his support. You can see the slides (with summarised findings and implications for Wikimedia) and watch a recording of the seminar linked from Wikimedia Australia's page about the event here:
https://lnkd.in/gJvhahii
|
|
Scooped by
Gilbert C FAURE
November 18, 7:03 AM
|
📢 Un appel à contributions pour un évènement annuel structurant sur le thème de l'éducation ouverte dans l'enseignement supérieur francophone.
N'hésitez pas à le relayer à vos réseaux si intéressés !
France Université Numérique L'Université Numérique Fondation UNIT Université Ouverte des Humanités IUT en Ligne AUNEGE UVED UNESS - Université Numérique en santé et en sport
|
|
Scooped by
Gilbert C FAURE
November 17, 5:02 AM
|
"YOU CAN'T GENERALIZE FROM A SINGLE CASE"
I hear this all the time. It’s wrong.
Galileo didn’t drop a thousand objects from a thousand towers. He used one case, a thought experiment. It was enough to kill Aristotle’s law of falling bodies, which had stood for 2,000 years.
One good case can beat a big sample.
That’s one of five myths I take apart in my paper “Five Misunderstandings about Case-Study Research.”
The paper just passed 31,600 citations on Google Scholar. That makes it the most cited paper in the 30-year history of Qualitative Inquiry. I never dreamed of this when I wrote it. Not even close. Thank you to all who made it happen 🙏
Now I’m trying something new.
arXivisual (a new platform for presenting papers) puts the full text next to short videos that walk you through it. You don’t just read the ideas. You see them.
Watch it here: https://lnkd.in/ebPfq99T
Free PDF of the full paper here: https://lnkd.in/e6amjYHu
Comments very welcome, kindly help share 🙏| 38 commentaires sur LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 17, 3:31 AM
|
IA et médecin : compte rendu, agenda, aide à la décision. Ce que l'IA fait au cabinet, ses limites et le cadre à respecter pour protéger vos patients.
Autre piège, plus discret, l’usage non maîtrisé d’outils grand public. Le projet de guide cite le CHU de Nancy, qui a décompté 440 000 connexions à des IA génératives personnelles par ses agents en août 2025.
|
|
Scooped by
Gilbert C FAURE
November 16, 4:30 AM
|
Eye tracking helps simulation teams compare how experienced clinicians and learners visually approach the same task, procedure or decision point, revealing differences that may be difficult to capture through observation alone.
|
|
Scooped by
Gilbert C FAURE
Today, 9:01 AM
|
Global Media and Information Literacy Week 2026 commemorations all around the world.
MILMedS (Media and information Literacies for Medical Students)
City/Country: Vandoeuvre lès Nancy, France
Date: 19 November 2026 – 31 Janurary 2027
Description: In the context of Master UE 7.309 (Méthodologies de la Recherche Médicale et Chirurgicale), medical students (french and chinese fromWuhan) in Medical Faculty/Université Lorraine will benefit from a conference on scientific and societal information tools in 2026, and will be involved in an active/self learning personal project to improve their Health and Media Information Literacy.
Organizer: Université Lorraine/CREM
|
|
Scooped by
Gilbert C FAURE
Today, 5:39 AM
|
Après plusieurs enquêtes de Next, huit médias ou sous domaines verticaux de médias viennent de faire l'objet d'une pénalité manuelle rarissime pour de grands éditeurs : la désindexation. Quand on tape Marie France dans Google, le site du magazine n'apparait plus. Idem pour Pleine vie, modes et travaux, Trucmania, Sciencepost et quelques autres.
Retour sur le calendrier, en quatre temps :
- Mi octobre, notre journaliste Jean-Marc Manach contacte Google au sujet de sites ou sous domaines émanant de grands éditeurs (Reworld, Ouest France, 20 minutes) dont il constate la très forte présence dans Discover, alors que les contenus sont manifestement générés par IA sans supervision humaine,
- début novembre, Google pénalise les sites concernés dans Discover, principal apporteur de trafic des sites d'actu,
- intrigués par la progression des audiences de Reworld alors que tous les médias d'actu reculent, nous publions une nouvelle enquête. Elle relie ce phénomène à une explosion du volume de publication, découlant d'un recours accru à l'IA, sur les conseils d'un référenceur bien connu du monde SEO,
- le 18 novembre, Google retire ces sites de son index.
Difficile d’établir un lien de cause à effet, Google ne détaille pas ses raisons et ne répond pas à nos sollicitations. C'est un coup dur pour les éditeurs concernés, qui risquent de voir leur trafic s'effondrer, mais un signal encourageant pour le web à moyen terme : Google semble enfin prêt à agir contre ceux qui s’appuient sur la notoriété de leur marque pour saturer son index et Discover de contenus au rabais, au détriment d’informations plus utiles.
L’IA en soi n’est pas le sujet : elle peut être employée à bon escient. Ce que nous cherchons à documenter, c’est la façon dont le volume décide de la visibilité, plutôt que la qualité.
Reste un paradoxe : ce même Google absorbe déjà une partie de l’audience des médias qui jouent le jeu avec ses Aperçus IA. Ce ménage est donc un bon début, mais certainement pas la fin de l'histoire. | 17 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 20, 9:55 AM
|
During the weekend, I had a fierce online discussion about hyper-productivity in academia and found how difficult it is to convince people that hyper-productivity is not something to be proud of or a measure of academic success. On the contrary, sometimes, too much is a sign of low quality. In Italian we say 'Il troppo storpia". I thus ended up reflecting on this issue and decided to write a bunch of things that I have a sense of urgency to communicate externally.
It resulted in this brief thought provoking article. I hope you will enjoy the reading.| 18 commentaires sur LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 20, 5:19 AM
|
Quand on pense connexion internet, on pense automatiquement à une box. Pourtant, ce n'est pas la seule solution pour accéder au web. 😉
Que vous soyez en déplacement, sur un salon pro, sur un chantier ou dans une zone où la fibre n'est pas disponible, plusieurs alternatives permettent de rester connecté en toute simplicité.
Dans cette newsletter, découvrez comment avoir internet sans box et quelle solution choisir selon vos usages pros. 💡
|
|
Scooped by
Gilbert C FAURE
November 19, 3:58 AM
|
🤖 L’IA change vite d’avis face à la vérité scientifique
🔎 Une étude montre que des modèles d’IA peuvent modifier leur réponse sur des faits scientifiques établis selon la formulation du prompt. Autrement dit, ils ne défendent pas toujours la même position sur la vérité scientifique, même lorsqu’elle est bien documentée.
🧪 Ce constat relance les questions de fiabilité de ces outils dans les domaines où la rigueur est essentielle, notamment en santé. Derrière leur assurance apparente, certains modèles se montrent sensibles au contexte, au cadrage de la question et à la manière dont on les sollicite.
Découvrez l’article complet sur Le journal du Médecin : https://lnkd.in/dZHva7cc
|
|
Scooped by
Gilbert C FAURE
November 19, 3:57 AM
|
I think the 1,135 likely fabricated references my team found in 500,000 science and engineering papers are a small part of the real number, and I said so when The Scientist interviewed me about our October PNAS study.
We examined journal papers in physics, chemistry, materials science, engineering and computer science, and we flagged references to studies that could not be found in any database. I am more concerned about the trend than the total, because the share of affected papers almost quadrupled between 2023 and early 2026. Yian Yin, a Cornell computer scientist who was not involved in the study, told the magazine that the scientific literature works as an ecosystem, so a problem that starts in a very small fraction of papers can have a larger effect over time.
We did not test whether AI writing tools caused the rise, and we suspect them because the timing matches (in a separate study, we found that many of these tools still invent citations). Our method checked only the references that came with a permanent web link to the cited source. My team is now examining the references without that link, and I expect to find a much larger share of the fabricated ones there.
Thanks to amazing Shelby Bradford for the interview. | 10 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 19, 3:50 AM
|
Elsevier, one of the world’s biggest academic publishers, asked 1,100 librarians in universities and colleges across 88 countries about their future. The most useful part of its new report, Librarian of the Future, comes from somebody else: the 19 researchers and academic leaders it interviewed. The report’s conclusion from those interviews is that researchers’ problem with libraries is more often not knowing what’s on offer, or uneven service, than a low opinion. Researchers at well-funded institutions valued help they barely noticed; others talked about missing databases and slow delivery of articles. One of the report’s main remedies is more storytelling about the library’s value, which won’t fix a missing database.
The headline numbers need a closer look. The claim that only 11% of librarians feel fully able to show their impact comes from a question asking what most holds them back, where being fully able was one choice on a list of obstacles and respondents could pick only one. The biggest skills gap, 27 points in negotiation, compares librarians’ own rating of their skill today with how important they expect it to be in five years, and 14 of those points come from expecting it to matter more. The note that should explain how skill levels were measured still contains unfilled placeholder text. Only 29 of the 1,100 librarians surveyed work in Australia. A third of all respondents named publisher pricing among the three biggest forces that will change research over the next five years, and the report gives it one sentence. It says nothing about the licence terms for using research articles with AI, or whose articles the new AI research tools draw on.
Librarians rated ethical judgement as one of their two most important skills, yet when ethical concerns stop some of them using AI, the report calls those concerns “a significant barrier to adoption”. If librarians are going to help oversee how AI is used in research, as 76% of them say they should, that role has to include the right to say no.
https://lnkd.in/gR4RZf3S
|
|
Scooped by
Gilbert C FAURE
November 18, 7:22 AM
|
Academic publishers then vs. academic publishers now.
The top figure is the 400-year-old Elsevier logo, intended to symbolize the symbiotic relationship between academic publishing companies and scholars.
The scholar nurtured the tree and picked fruits with his hand, but he did not realize that the tree would grow and the fruits would be out of his reach without a ladder (subscription).
Image from "From symbiont to parasite: the evolution of for-profit science publishing" by Peter Walter and Dyche Mullins. | 21 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 18, 4:51 AM
|
Le World Economic Forum vient de publier un rapport sur l’avenir de l’intégrité de l’information, et l’un de ses constats concerne directement ceux qui font de la veille leur métier. À mesure que l’information passe par des assistants d’IA, la recherche cède la place à des réponses directes. Le confort est réel, mais une synthèse unique peut masquer les preuves, l’incertitude et les désaccords qui entourent une question. Le rapport cite une évaluation de l’EBU et de la BBC qui a relevé des problèmes importants d’exactitude ou de sources dans 45 % des réponses testées.
C’est une question qui nous a accompagnés tout au long de la conception de Curebot Assistant, pensé pour amplifier le travail des analystes plutôt que pour s’y substituer. Le rapport la pose à l’échelle de tout l’écosystème : même les systèmes qui s’appuient sur des bases documentaires vérifiées ne garantissent ni la qualité ni l’exhaustivité des sources, et ils exigent une gouvernance des sources, une vérification des citations et une relecture humaine dès que l’enjeu est important.
Plus l’information nous parvient sous forme de réponses synthétiques, plus il devient important d’en garder la trace : savoir d’où vient chaque élément, pouvoir remonter à la source, confronter plusieurs lectures avant de transmettre une analyse, et laisser à l’humain la responsabilité du jugement. Ces exigences ont toujours fait partie du métier de veilleur, mais elles deviennent la condition pour que l’IA fasse gagner du temps sans faire perdre en fiabilité.
À l’échelle d’une organisation, la fiabilité d’une analyse tient à la façon dont veilleurs, experts métiers et décideurs vérifient et discutent ensemble ce qui leur parvient.
Le rapport du World Economic Forum, The Future of Information Integrity 👉 https://lnkd.in/eVJzagXf »
Christophe Deschamps, Terry ZIMMER, Claude Revel, Charles PAHLAWAN, Anne-Solène SPIDO, Ludovic Huth, Clémence Jost, Laïdi Ali, Souveraine Tech
|
|
Scooped by
Gilbert C FAURE
November 17, 5:02 AM
|
Stop Asking Claude:
“Find a Research Gap.”
Use these 16 prompts to analyze research papers, connect ideas, and find hidden insights.
(Save this now 🔖 Research Faster and Smarter)
Add this instruction to any prompt you use: “For each evidence claim, cite the paper and its page, section, table, or figure. Flag inferences and missing details. Describe gaps within the papers I supplied.”
1. Map what the papers actually tested
For [research question], make a table with one row per study. Record the sample or dataset, setting, design, comparison, outcome, and main result. Use the table to identify one missing comparison that would help answer the question. Explain why that comparison matters.
2. Explain why two papers disagree
Compare how [paper A] and [paper B] studied [finding]. Check whether their samples, definitions, methods, and timing are comparable. Identify up to two differences that could explain the disagreement. Propose one analysis or experiment that would help distinguish between those explanations.
3. Check who the evidence applies to
I want to apply findings about [topic] to [target population or setting]. Compare that target with the studies’ eligibility criteria, actual samples, and recruitment settings. Identify the biggest mismatch and turn it into a question about whether the finding holds in my target group.
4. Find where an effect might change
For [observed effect], list the doses, intensities, or exposure levels the studies actually tested. Pick one untested range where the results give a reason to expect a change. Explain that reasoning and specify the two levels a follow-up study should compare.
5. Find the missing time point
For [outcome], record when each study measured it relative to [event, exposure, or intervention]. Identify one missing time point that would help distinguish a temporary change from a lasting one. State what you would expect to observe at that time under each explanation.
6. Test a proposed mechanism
The papers suggest that [X] affects [Y] through [mechanism]. Map the proposed chain and label which links were measured directly, tested experimentally, or inferred. Pick the least supported link and propose one measurement or experiment that would help test it.
7. Spot a measurement mismatch
Compare how the papers measured [construct]. Extract the instrument or indicator, scoring rule, cutoff, and measurement window where reported. Identify one pair of results that could look comparable despite measuring different things. Propose a study that applies both measures to the same sample.
More Prompts in comment below 👇
|
|
Scooped by
Gilbert C FAURE
November 16, 1:21 PM
|
12th Annual Beyond Sciences Initiative (BSI) Conference_FREE REGISTRATION_ 26–27 February 2027 |
Six outstanding international keynote speakers:
Megan Sykes — Columbia University, USA
Eduardo J. Villablanca — Karolinska University Hospital, Sweden
Elina Zuniga — UC San Diego, USA
Magdalena Plebanski — RMIT University, Australia
Fabiola Osorio — University of Chile, Chile
Sourav Ghosh — University of Minnesota, USA
SHOWCASE YOUR RESEARCH WITH US: we welcome abstract submissions from researchers and trainees across infectious diseases, cancer, chronic diseases, global health, biotechnology and bioinformatics.
Abstract submission deadline: 1 December 2026
Register for free: https://lnkd.in/gJjUm4zW
#BeyondSciencesInitiative #ScientificConference #Research #EarlyCareerResearchers #GlobalScience
|
|
Scooped by
Gilbert C FAURE
November 16, 5:05 AM
|
Update: I tested three more AI-powered tools for literature search, bringing the total number tested to 15. Two new ones were really bad, one finding only one and the other literally none of the studies in the target sample. One performed somewhat better but still rather poorly. I have updated the paper with the new results and a perspective on brighter horizons here https://lnkd.in/ehnmpDcT. I will present the paper at the Leibniz Association Open Science day in Berlin on Tuesday. After the conference I plan to submit the paper for review in Open Research Europe, the new publish - review - curate platform for research in Europe. Your thoughts about the validation study are much appreciated. I particularly hope that the tool producers will review the paper and respond. If you are an employee at Anthropic, Google DeepMind, Undermind, PapersFlow, Elicit, Elsevier | Scopus or know someone working there, please alert them.
https://lnkd.in/ewfx7Rev | 15 comments on LinkedIn
|
|
Scooped by
Gilbert C FAURE
November 16, 4:24 AM
|
It is time for universities to stop reducing the assessment of colleagues to Scopus-based indicators, journal rankings and publication counts.
I write as someone the present system serves well. I have just published an article in a Q1 journal, I have one book in press with Routledge and another with Oxford University Press, and I have others under contract. I would rather colleagues assessed that work by what it argues and the evidence it brings than by the quartile of the journal or the name of the press.
The San Francisco Declaration on Research Assessment (DORA)(sfdora.org) offers a different approach. It came from within the scientific community. Journal editors and publishers drafted it at the 2012 annual meeting of the American Society for Cell Biology, and it was published in 2013.
Monographs, critical editions and articles in leading journals should remain central to academic life. Committees should assess what these works contribute: the originality of the argument, the quality of the evidence, the care taken over method and the significance for the field. The reputation of a publisher or the position of a journal in a ranking records where a work appeared, and a committee should treat it as context for its own reading of the work.
DORA also asks institutions to consider contributions other than conventional publications, including datasets and software, and to take account of qualitative evidence such as influence on policy and practice.
For those of us in the humanities, this should include scholarly translations, archival and editorial projects, digital resources, exhibitions and sustained public scholarship. It should also include research published in languages other than English, and work whose significance becomes apparent over years rather than within a short assessment cycle.
Outputs will still differ in weight. Departments should set clear criteria suited to each discipline, and the colleagues who apply them should take responsibility for an informed evaluation. DORA's guidance supports the responsible use of indicators in context, as one source of information among several.
Many leading European universities have already signed. In the United Kingdom they include Oxford, Cambridge, UCL, King’s College London, Edinburgh and Manchester. In the United States, to my surprise, fewer research universities have signed, and some signatories are units within a university: an appointments and promotions committee at the University of Chicago and a doctoral programme at the CUNY Graduate Center. In the Arab world I know of only one university that has signed, the American University of Sharjah.
Universities should sign DORA. More importantly, they should apply its principles in recruitment, annual reviews, promotion and tenure.
#DORA #ResearchAssessment #ResponsibleMetrics #HigherEducation #Humanities #AcademicPublishing #PromotionAndTenure #AcademicAssessment #Assessment #Universities #Promotion | 10 comments on LinkedIn
Your new post is loading...