Randomized experiment:
Using a LLM (ChatGPT 5 Instant) in a Wikipedia editing assignment improved undergraduates' "linguistic polish", but had no significant effect on content quality and verifiability (i.e. citation coverage and reliability of sources) https://t.co/BKQBVemJKC
via https://twitter.com/WikiResearch/status/2064865965196857425
Large Language Models on Wikipedia Editing: The Stylistic Mask Effect in a Randomized Controlled Intervention in Health Education
The effects of large language models (LLMs) on content, references, and language quality in Wikipedia assignments remain unclear. This randomized controlled study assessed how model assistance influences the quality of Wikipedia edits created by undergraduate audiology students. Thirty-six participants were assigned to two groups: Group 1 (G1) edited without model support, and Group 2 (G2) edited with ChatGPT support. Twenty blinded expert reviewers evaluated a sample of 30 texts (15 per group) using a six-item Likert-scale instrument covering Content, References, and Language. Inter-rater reliability was high (Gwet's AC2 = 0.80). The analysis suggests that LLM assistance may lead to a significant improvement in Language (β = 0.400; p = 0.023; Cliff's δ = 0.400). G2 improved by 0.68 ± 0.47, while G1 improved by 0.28 ± 0.57. The analysis found no significant differences for Content (β = 0.167; p = 0.214) or References (β = −0.267; p = 0.859), although G2 scores in the latter category trended lower. Output counts were not associated with quality improvement in either group. This asymmetry is proposed as a potential Stylistic Mask Effect. It describes a gap between surface language gains and limited substantive development. LLM assistance appears to risk substituting constructive contribution with linguistic polish. This record contains the preprint manuscript. The manuscript has been submitted for peer review and has not yet been published. Dataset and code are made available to support transparency of the reported analyses.
"Multilingual and Cross-Lingual Citation Needed Detection on Wikipedia for Lower-Resource Languages" https://t.co/z0p1oyT6AU (Data and code: https://t.co/jjSgdde5LP )
"SLMs [small language models] outperform LLMs on monolingual CND" https://t.co/TncDx3l8p4
via https://twitter.com/WikiResearch/status/2064593531768029508
"Institutionalization and automation of [newcomer adaption] processes in Wikimedia projects" https://t.co/alVEkW9Gcm
"A clear correlation was observed between the number of users registering and the number of people actively editing on Wikipedia." https://t.co/oQCJg1lJue
via https://twitter.com/WikiResearch/status/2064424609475403854
"Orthographic Practices in the Alemannic Wikipedia" https://t.co/9eMVxW2cDS
(thesis, finds that the project "promotes broad and inclusive norms and guidelines that allow for multiple, diverse orthographic practices" for the German dialects it covers) https://t.co/X9uZduifL3
via https://twitter.com/WikiResearch/status/2063471371720818737
"Designing for Human–AI Collaboration in Open Knowledge Work":
Lessons from the failure of the Wikimedia Foundation's "Computer Aided Tagging" tool on Wikimedia Commons https://t.co/yfSCidpsLc https://t.co/xapeyhNwbX
via https://twitter.com/WikiResearch/status/2062945253267841235
Computer-Aided Tagging on Wikimedia Commons: Designing for Human–AI Collaboration in Open Knowledge Work
RT @CristianCantoro: Honored to have received the this award for our demo "Wikipedia Community Health Dashboards", check them out here:
htt…
via https://twitter.com/WikiResearch/status/2062069841062981720
WikiResearch (@WikiResearch) on X
RT @CristianCantoro: Honored to have received the this award for our demo "Wikipedia Community Health Dashboards", check them out here:
htt…
"Monitoring the gender gap in the coverage of biology professors on Wikipedia" https://t.co/bqjo1WaiVS
Women "were significantly less likely [to be on WP] until 2018. However,the trend reversed in 2022,and women were 25% more likely than men to have a Wikipedia biography in 2024" https://t.co/Ciao3KoUJt
via https://twitter.com/WikiResearch/status/2062065516269978045
In the latest issue of our newsletter:
* "Wikilambda the ultimate: the Wikimedia foundation’s search for the perfect language" (with Wikifunctions/Abstract Wikipedia)
And other recent research publications involving Wikidata and Wikipedia
https://t.co/qPKXQR4hgo https://t.co/zIRbLdhC16
via https://twitter.com/WikiResearch/status/2061853022729429326
Research:Newsletter/2026/May - Meta-Wiki
@Wikipedia @JemielniakD "@Wikipedia at 25: The Encyclopedia That Might Not Last" @JemielniakD follows up on how the extraction of its value by artificial intelligence and other factors might break Wikipedia.
https://t.co/Lhm6twrW9p
via https://twitter.com/WikiResearch/status/2056676391807025394
"@Wikipedia at 25: The Encyclopedia That Shouldn't Exist" @JemielniakD on how its governance structure, quality control mechanisms, and integration into educational practice make Wikipedia work.
https://t.co/REbUUZYnuI
via https://twitter.com/WikiResearch/status/2056674164270944290
