WikiJournals are an experiment applying scholarly peer review to improve Wikimedia content. From professors to students, all authors are treated the same. Their work is sent out to experts in the topic for comments, criticism and recommendations. Articles that pass through this process are published as citable 'versions of record', assigned a DOI, formatted up into a PDF as a permanent read-only version separate from any associated Wikipedia article, and indexed like articles in other science journals in the major scholarly databases.

Started in 2012 with the medical journal (WikiJMed), the group has expanded to include two additional sister journals that cover science, technology, engineering, and mathematics topics (WikiJSci) and Humanities, Arts, and Social Sciences topics (WikiJHum). WikiJournals were profiled in The Signpost back in 2016 and in 2017, with similar projects mentioned in 2012 and in 2014.

Many WikiJournal articles become Wikipedia pages - some written from scratch (e.g. a two-sentence stub expanded into a full article), and some submitted straight from Wikipedia (e.g. Rosetta Stone). As well as these articles that can form Wikipedia pages (or sections of pages, e.g. table of pediatric medical conditions and findings named after foods), the WikiJournals have also published original research (e.g. A card game for Bell's theorem and its loopholes), medical case studies (e.g. acute gastrointestinal bleeding from a chronic cause: a teaching case report), images and galleries (e.g. cell disassembly during apoptosis and medical gallery of Blausen Medical).

The journals use a radical form of academic publishing - they are free both for readers and for authors (an extreme rarity in the academic world). Other journals are paid for either by their readers, who subscribe to a journal or belong to an institution like a university or a library which pays a subscription, or by their authors, who are obliged by open access journals to pay for the cost of review and publication. With the WikiJournal model, open access publishing becomes affordable for all authors, including researchers from less wealthy countries, early career researchers, and even students.

Not only that, but parts of the academic publishing process that are usually done behind closed doors are also published openly, such as peer reviewer comments and author responses. Indeed, more than 75% of WikiJournal peer reviewers choose to forgo anonymity, reviewing openly under their names. Even much of the journal's organisation and planning is done publicly: for example the assessment of whether the journals will be Plan S-compliant (here), and indexing applications for Free Journal Network (here), ASAPBio ReimagineReview (here), or SCOPUS (here).

The WikiJournal format is a good way of getting outside expert feedback on existing Wikimedia content (as peer reviewers), as well as attracting contributions from academics who'd otherwise have been unlikely to contribute (as authors). At scale, a more developed platform could even improve knowledge equity by lowering the cost and technical barriers to academic publishing.

Within the wiki ecosystem, these have a number of unusual features. Firstly, peer reviewers are specially contacted based on their expertise, so often it is their first ever contribution to a Wikimedia project. There is also a large amount of partnership with outside organisations. The journals' publication ethics statement has been audited by the Committee on Publication Ethics. There is also thorough vetting in order to be added to included in academic indexes (e.g. the Directory of Open Access Journals).

The WikiJournal model aims to be complementary to Wikipedia's existing Featured article and Good article review systems. These systems involve review by an editor (in the case of GA) or 5-15 editors (in the case of FA) who are not necessarily subject-specialists, but implement extensive criteria checklists to assess citations, comprehensiveness, and readability. A WikiJournal review is more formal, in that the journal editors invite external experts in the field, who may review anonymously if they wish, and the decision to accept is made by the editors on the basis of those reviewers' comments and author responses. Another major difference is that the reviewed article is then stable; its text may well be used to update the corresponding Wikipedia article, but even if the Wikipedia article drifts away from the reviewed version, the journal article's text remains as the reviewed, approved and citable WikiJournal version. New versions can be made but have to remain as preprints until they've been peer reviewed and assigned an updated DOI. This may be particularly useful for attracting additional contributors for some of the high/top importance stubs, articles in need of expert attention, or articles in need of rewriting.

2018/19 has seen particular growth, with the creation of WikiJSci and WikiJHum and expansion of WikiJMed to bring the joint editorial boards up to 63. In 2018, of the 27 articles submitted, 9 were published, 3 were declined, and 1 was withdrawn from consideration. The turnaround time still varies pretty widely, but is improving. Readership from DOI-clicks is reaching levels typical of academic journals (800 per article) and is increasing. Journal articles that have been incorporated into Wikipedia unsurprisingly dwarf this readership, at 4.2 million during 2018 (WikiJMed, WikiJSci, WikiJHum).

So what're the plans for the future? Well, one major question is whether it is beneficial for the initiative to become a full sister project. Please take a look and comment in the discussion section.

Proposal: WikiJournals as a new sister project

This would entail new technical features better suited to the journals' workflow, their own dedicated website, and increased visibility and support from the WMF.

The journals also aim to stay agile and experimental, looking at new ways to tackle peer review, assessment, community engagement and integration with other Wikimedia projects. Perhaps closer partnerships can be formed with existing academic journals to help find reviewers. The platform could even compete with non-free platforms like Google Scholar and ResearchGate by integrating multiple free services like Scholia academic profiles, Wikicite's bibliographic indexing, the machine reading of ContentMine or hosting for author self-archived 'green open access' articles through commons/wikisource.