Littview @littview.bsky.social · 10hMy guess is the slightly lower counts are just a lagged S2 snapshot. The OpenAlex column is the fun part though, 2 vs 68 on one paper. Doesn’t matter much for discovery, but for citation chasing in a review you’d want to say which index you used. 000
Littview @littview.bsky.social · 12/10/2026 Same pattern here. A vague question just gets you a fast, vague review, whether people or AI do the screening. Teams that pilot their extraction form on a few papers first tend to find the holes in their question early, from what we’ve seen. Cheap step too. 000
Littview @littview.bsky.social · 08/10/2026Agree, and from the outside it still looks like a long literature search. The real skill is in screening and appraisal, the judgment calls nobody sees in the final paper. 000
Littview @littview.bsky.social · 08/10/2026For what it's worth, the hesitation we hear about isn't AI itself, it's tools that don't show how they decided. Agreement stats against a human reviewer and a log of every exclusion seem to do more for adoption than any accuracy claim. 000
Littview @littview.bsky.social · 04/10/2026if the order past a point is basically work ID, recall@N on the plain OQL run isn't really testing ranking. So why does proximity jump it to 93.9%? Dropping 1.5k records alone doesn't seem enough. Does proximity switch on relevance scoring? 100
Littview @littview.bsky.social · 04/10/2026Full-text screening next, references in Zotero, no PDFs? Our free plugin: select, right-click, and it attaches legal open-access copies via OpenAlex, by DOI, PMID, arXiv ID or title. www.littview.com/blog/zotero-... #SystematicReview #EvidenceSynthesis #AcademicSkylittview.comA Free Zotero Plugin to Find Missing PDFs | LittviewOpenAlex PDF Finder is a free Zotero plugin that attaches legal open-access PDFs to the references you select, by DOI, PMID, arXiv ID or exact title. 040
Littview @littview.bsky.social · 03/10/2026Covidence, Rayyan, Zotero, Elicit, NVivo: not competing answers, different stages of a review. A map of 24 literature review tools across 8 stages, with the overlaps shown. www.littview.com/blog/literat... #SystematicReview #EvidenceSynthesis #AcademicSkylittview.comLiterature Review Software by Stage | LittviewWhich literature review software does what, stage by stage: PubMed, Zotero, Covidence, Rayyan, Elicit, NVivo, VOSviewer and more, with the overlaps shown. 061
Littview @littview.bsky.social · 02/10/2026The surprising part for me isn't the $1.46, it's the one-question setup beating the five-question one. Simpler prompt, better recall. Humans still own the borderline calls though. 000
Littview @littview.bsky.social · 30/09/2026Full disclosure, we build SR software that uses AI. Some steps do get faster: dedup, a first screening pass you then check against a human-screened sample, the PRISMA counts. That's real time saved.... 000
Littview @littview.bsky.social · 30/09/2026But the protocol, the search strategy, inclusion calls, risk of bias, the synthesis, that's where the time goes and where the judgment is. A weekend review has to have cut some of that. I'd want to know which parts. 000
Littview @littview.bsky.social · 28/09/2026Gusenbauer & Haddaway, right? Good one to send anyone planning to run their review off Google Scholar alone. Only about half of the 28 held up as principal search systems, and GS wasn't one of them. 010
Littview @littview.bsky.social · 28/09/2026Worth adding: finding and extracting is one part. A systematic review also needs dual screening, conflict resolution and PRISMA counts you can defend. We're building Littview to bring AI into that part too, with humans making the calls (full disclosure, it's ours). Nice to see the space growing. 010
Littview @littview.bsky.social · 26/09/2026Congrats on v4! As people building screening tools, RAISE 2 is the part we keep going back to. Having a clear standard for how to test and report these tools makes our job a lot easier. Will dig into the new version. 000
Littview @littview.bsky.social · 24/09/2026The AI isn't the problem, the reporting is. RAISE asks that any AI use which makes or suggests judgements be fully and transparently reported, with human oversight. Thematic synthesis is judgement, so "verified by a researcher" needs to say what was checked and what changed.pmc.ncbi.nlm.nih.govPosition Statement on Artificial Intelligence (AI) Use in Evidence Synthesis Across Cochrane, the Campbell Collaboration, JBI, and the Collaboration for Environmental Evidence 2025Evidence synthesists are ultimately responsible for their evidence synthesis, including the decision to use artificial intelligence (AI) and automation, and to ensure adherence to legal and ethical st... 100
Littview @littview.bsky.social · 23/09/2026Writing inclusion and exclusion criteria? Four that most reviews need, two that depend on your question, and what to leave out. Examples from education, HR, supply chain and AI. #SystematicReview #EvidenceSynthesis #AcademicSkylittview.comInclusion and Exclusion Criteria Examples | LittviewInclusion and exclusion criteria examples for a systematic review in any field: the criteria to write, what to leave out, and how to apply them. 021
Littview @littview.bsky.social · 22/09/2026I think excluding outright would throw away too much, but the gap is real. Lundh's 2017 Cochrane review found industry-sponsored drug and device trials had more favourable results, and standard risk of bias assessment didn't explain the difference. Cochrane is piloting a tool for it, TACIT. 010
Littview @littview.bsky.social · 19/09/2026LLMs should do the mechanical work, humans make the judgement call. Final decision should always be human. 010
Littview @littview.bsky.social · 18/09/2026Conference session title: "I'm Not Gonna Do It: Filling in the PRISMA 2020 Diagram by Hand."static.klipy.comLoretta Scott Says 'I'm Not Gonna Do It'ALT: Loretta Scott Says 'I'm Not Gonna Do It' 010
Littview @littview.bsky.social · 16/09/2026Their section 5 checklist should just become a required box to tick before submission. Would save reviewers so much guessing. 110
Littview @littview.bsky.social · 08/09/2026Two minutes on how a systematic review runs in Littview: from the search export to a Results section you can defend, with AI beside you at every stage. www.youtube.com/watch?v=mFLf... #SystematicReview #EvidenceSynthesis #AcademicSkyyoutube.comLittview: Systematic reviews, end to end (2-minute intro)YouTube video by Littview 000
Littview @littview.bsky.social · 06/09/2026Which critical appraisal tool for which study design? One map: RoB 2, ROBINS-I, Newcastle-Ottawa, CASP, JBI, MMAT, EPHPP, QUADAS-3, AMSTAR 2, ROBIS. What each one gives you back, and every official checklist linked. www.littview.com/blog/critica...littview.comCritical Appraisal Tools for Systematic Reviews | LittviewWhich critical appraisal tool fits which study design: RoB 2, ROBINS-I, Newcastle-Ottawa, CASP, JBI, MMAT, QUADAS-3, AMSTAR 2, ROBIS. Checklists linked. 010
Littview @littview.bsky.social · 03/09/2026Littview is systematic review software with AI assistance. Import, screen, extract, and report your review in one place. AI assists at every step, and you always make the final call. Free to start, no time limit: littview.com #SystematicReview #EvidenceSynthesis #AcademicSkylittview.comLittview — Systematic Review Software with AI AssistanceImport, screen, extract, and report your systematic literature review in one place. AI assists at every step — you always make the final call. 010
Littview @littview.bsky.social · 03/09/2026Choosing review software comes down to one question: which stages of your review does it cover, and how? Covidence, Rayyan, Elicit, and Littview compared, from each vendor's own documentation. #SystematicReview #AcademicSkylittview.comCovidence vs. Rayyan vs. Elicit vs. Littview (2026) | LittviewA stage-by-stage feature comparison of four systematic review tools across import, screening, extraction, risk of bias, PRISMA, reporting, and pricing, from each vendor's own documentation, checked 3 ... 000