Talk:WikiJournal of Science

From Wikiversity
Jump to: navigation, search

Discussing the editorial guidelines[edit]

How many peer reviewers?[edit]

Currently the guidelines say we need at least two. I would argue that as a general guideline a minimum of one reviewer is enough:

  • Finding reviewers is not easy.
  • Filtering an abundance of weak contributions may not be an important problem, at least initially.
  • In some fields (high-energy physics) and some journals (JHEP), one reviewer is the norm.

Sylvain Ribault (discusscontribs) 13:18, 19 November 2017 (UTC)

I tend to agree. Boris Tsirelson (discusscontribs) 13:42, 19 November 2017 (UTC)

re

I'm a little reticent about reducing down to just a single peer reviewer since I think its' worth erring on the side of caution. However, I realise that I'm biased by the norms of my field (biomed) where 2-3 reviewers is the norm. I think a case could be made for it depending on how in-depth the review is. I think it will be important to have the general consensus of the board and other editors on this issue, so I'll send an email to the mailing list highlighting it. T.Shafee(Evo﹠Evo)talk 10:57, 21 November 2017 (UTC)
I agree, 2 would be good to be safe and more may be more difficult to get. --Joanna Argasinska (discusscontribs) 11:13, 21 November 2017 (UTC)
As part of the discussion, here are a couple of counterpoints on reviewer number:
  • Particularly whilst the journal is starting out, being thorough may be prioritised over being rapid (however taking too long can definitely put off authors)
  • Broad encyclopedic articles can benefit from reviewers from multiple backgrounds (though this will not nbe true of all articles)
  • Sometimes reviewers can have very different opinions on an article, and so having more than one reduces stochastic variation (though whether this is sufficient to make a difference is up for debate)
T.Shafee(Evo﹠Evo)talk 11:22, 21 November 2017 (UTC)
How about we make one reviewer the default and leave it to the editor's discretion to involve a second reviewer (in particular if the first review was very short, and didn't have many specifics). Markus Pössel (discusscontribs) 11:56, 21 November 2017 (UTC)
We are not talking about desirable or best practices here, we are talking about a rule that should be obeyed in all cases for all fields. For the sakes of the quality of articles and reputation of the journal, I would trust the editors to do as they see fit on a case by case basis, and I would expect them to require at least two reviewers in most cases. The only reason that I see to set a limit on the number of reviewers is for the journal to officially qualify as a peer-reviewed publication. For this, a limit of one reviewer is enough. Having a higher limit just restricts the editors' options, for no tangible benefit. And given how widely practices differ between fields, I am not sure that it is feasible to have a default option as Markus Pössel suggests. Sylvain Ribault (discusscontribs) 12:06, 21 November 2017 (UTC)
I think that we should set the standard of 2 peer reviewers and not one. It is important to maintain the standards of a standard journal and we should not compromise at least while setting norms. I understand that it might be hard to get reviewers. But it is also important to ensure quality. I think that the purpose does not get served if there is a single reviewer. We can review this norm after based on the hardship we face in the course of processing the next five articles that are submitted. Diptanshu💬 12:40, 21 November 2017 (UTC)
I think that we need to have at least two reviewers for assuring quality of content. In clinical neurorehabilitation where I am doing my PhD, there would at least be three reviewers. I think two reviewers should be the default, and the editor can have the discretion to reduce or increase the number of reviewers depending upon the type of the article and the background of the chosen reviewers. --Netha Hussain (discusscontribs) 13:00, 21 November 2017 (UTC)
I am surprised we are even thinking about allowing only one. If experts are hard to find in a field, we can always have one expert and one generalist who can comment on structure, English, comprehensibility, illustration, copyright, and layout. I would like to see three reviewers for everything but could accept a minimum of two if we're really in difficulty. Allowing only one opens us up to charges of bias and unprofessionalism, hardly a great way to start. Chiswick Chap (discusscontribs) 13:12, 21 November 2017 (UTC)
I peer-review for some high impact academic journals in my field they only require one these days. I do comment on general issues too though so I agree with Chiswick Chap. It is going to be hard to find more than one disciplinary expert to per review. A second peer reviewer should speak to the general issues. Fransplace (discusscontribs) 20:18, 21 November 2017 (UTC)
A single reviewer strikes me as dangerous. I wasn't aware that was the norm in some fields and/or journals, but seeing how widely opinions re suitability can diverge even with some of the entirely non-controversial papers I have been involved in reviewing (prevalance of tail streamer asynchrony in frigate birds!...), I don't think it's a safe bet. Neither will it do any good for eventual acceptance of the journal as a reliable source. Three may be tricky to do, but I think we should shoot for a minimum of two. --Florian (Elmidae) (talk · contribs) 15:32, 21 November 2017 (UTC)
I also feel that one reviewer would be very dangerous, given the variability between reviewers, and would give an impression of low standards in the large fields where 2 or 3 is the norm. What can be done when it is difficult to find reviewers is to find one reviewer who is totally independent of the editor, and complete with a close colleague or student of the editor, or a review by the editor her/himself. I have done this (asking a student of my lab as 2nd reviewer) on occasion for traditional journal articles, and have performed the 2nd review myself in very rare cases.Marcrr (discusscontribs) 09:32, 22 November 2017 (UTC)

I may be similarly biased from a biomedical standpoint and I am fully sensible to the fact that finding 2 expert peer reviewers can significantly prolong the peer review process and may even be impossible... I would also suggest to have 2 reviewers as default but leave it to the handling Editor's discretion, with or without back-up discussion with the editorial board, to allow other alternatives as mentioned here by others. -Steven Fruitsmaak (Reply) 20:13, 22 November 2017 (UTC)

I am rather neutral regarding whether we should have a minimum of 1 or 2 reviewers. The minimum of 1 means much easier and faster article processing, and still means that we can request additional reviewers if there is any doubt, and there will be additional quality assurance by readers after the work is published. On the other hand, a minimum of two reviewers means a more reputable journal, with a greater chance of being included in major scientific websites, which will increase readership and probably also participation. Mikael Häggström (discusscontribs) 18:45, 23 November 2017 (UTC)

To summarize, allowing one reviewer seems dangerous to many people. I propose to leave the guidelines unchanged for now with the "two external peer reviews", with the option of revisiting the issue later if need be as Diptanshu suggested. Sylvain Ribault (discusscontribs) 21:57, 27 November 2017 (UTC)

Message templates[edit]

Most templates are in the format .docx. In my professional life I have never used this format, and all messages to and from journals have been in plain text or Latex. The format .docx would look like a baroque archaism to colleagues in my field. Why not have plain text templates? OK this would eliminate the pretty logo, but then the template could be directly pasted in an email body. And the template would be more easily edited. Sylvain Ribault (discusscontribs) 13:26, 19 November 2017 (UTC)

I agree. Boris Tsirelson (discusscontribs) 13:38, 19 November 2017 (UTC)
We originally had .msg format to be directly sent from outlook, however we moved over to .docx so that the content could be copy-pasted into an email (not sent as attachment). Copy-pasting from from MSword typically retains basic formatting in e.g. outlook and gmail (also images and hyperlinks). We could also have a plaintext version if it is creating copy-paste artefacts, however it can be harder to keep the two versions synchronised if changes are made to the template. T.Shafee(Evo﹠Evo)talk 04:12, 20 November 2017 (UTC)
Outlook? .msg format? MSWord? What's that? Different communities use different tools, and some tools and formats you are mentioning are alien to me. It would be good to have a reference format for templates that would be universal, such as plain text, HTML, PDF. Also, it would be good to be able to change the template without having to upload a file somewhere. This would be possible if the template was part of a wiki page. This would not prevent people from uploading "secondary" templates in their preferred formats. Sylvain Ribault (discusscontribs) 20:14, 20 November 2017 (UTC)
Indeed, these formats (alien to you) are alien to me, too. Though I understand that they (at least some) are habitual for persons alien to TeX. Boris Tsirelson (discusscontribs) 20:38, 20 November 2017 (UTC)
That's fine by me, we can make a WikiJournal_of_Science/Editorial_guidelines/Invitation_1 subpage etc to house the plaintext email templates. T.Shafee(Evo﹠Evo)talk 10:44, 21 November 2017 (UTC)
I agree having them in wiki will be easier. Mikael Häggström (discusscontribs) 18:48, 23 November 2017 (UTC)
I propose to create one subpage for all message templates, rather than one subpage per template: WikiJournal_of_Science/Editorial_guidelines/Message_templates. This would make it easier to add new templates. And that subpage could then have some structure, and a table of contents. It seems that I cannot create such a page myself, could someone authorized please do it, as soon as there is an agreement on this? Sylvain Ribault (discusscontribs) 22:04, 24 November 2017 (UTC)
You cannot create such a page? Strange. I have no additional rights, but I just did it. We both are implicit members of: Autoconfirmed users. Boris Tsirelson (discusscontribs) 22:33, 24 November 2017 (UTC)
Thank you! Then it was my technical limitations rather than the lack of authorization. Sylvain Ribault (discusscontribs) 22:41, 24 November 2017 (UTC)
Agreed, a single page is a sensible structure. I'll copy in the text today and link over from the editorial guidelines. T.Shafee(Evo﹠Evo)talk 00:47, 25 November 2017 (UTC)

Original research[edit]

"Original research on any medical or biomedical topic can be submitted. Such papers follow the standard Introduction, Results, Discussion, Methods format, with any relevant ethics approval. Publication of supplementary data sets is encouraged." — No more medical, and probably not just this format. Boris Tsirelson (discusscontribs) 09:20, 25 November 2017 (UTC)

Thanks for spotting this! It was a copy and paste error from the Medical journal's guidelines. Should be corrected now. T.Shafee(Evo﹠Evo)talk 11:03, 25 November 2017 (UTC)

How to deal with epistemology and history of science?[edit]

In scientific texts, we often have some words on the history of the subject, if only to summarize previous works and say who did what. For example, in the historical section of the Wikipedia article on General relativity, we find the statement:

The Einstein field equations are nonlinear and very difficult to solve. Einstein used approximation methods in working out initial predictions of the theory. But as early as 1916, the astrophysicist Karl Schwarzschild found the first non-trivial exact solution to the Einstein field equations, the Schwarzschild metric.

On the other hand, some higher-level historical statements result from the works of historians of science, rather than scientists themselves. For example, in the historical section of the Wikipedia article on Quantum mechanics, we find the statement:

Following Max Planck’s solution in 1900 to the black-body radiation problem (reported 1859), Albert Einstein offered a quantum-based theory to explain the photoelectric effect (1905, reported 1887). Around 1900-1910, the atomic theory and the corpuscular theory of light first came to be widely accepted as scientific fact; these latter theories can be viewed as quantum theories of matter and electromagnetic radiation, respectively.

Similarly, some high-level interpretative statements are not about science itself, but about epistemology. The question is how to deal with such statements about epistemology or history of science. It is difficult to ask the usual editors and reviewers to assess these statements, since they are scientists, not historians or philosophers. I can see a number of possible attitudes, in increasing order of difficulty:

  1. Ignoring the issue as not very important.
  2. Asking authors to remove any statement that is not clearly scientific in the submitted texts.
  3. Recruiting an editor who is a historian of science.
  4. Coordinating with WikiJHum in order for them to assess the relevant content. Similarly there could be some coordination with WikiJMed for articles that involve both biology and medicine.

Sylvain Ribault (discusscontribs) 14:40, 27 November 2017 (UTC)

It would be my expectation that such statements could be assessed by subject-area expert reviewers; or if necessary, flagged by them as in need of further investigation. I.e., I'd think that a qualified reviewer of an article on quantum mechnics would be able to pass some judgement on a statement about the history of their field, or note statements that require sourcing. After all, such general introductory passages are not rare in scientific articles, and AFAIK journals do rely on the standard reviewing process to vet them along with the more specialized content. --Florian (Elmidae) (talk · contribs) 14:29, 27 November 2017 (UTC)

Tables of contents[edit]

Why does this submitted version of a Wikipedia article not have a table of contents? Shouldn't we at least have the option of toggling a table of contents? Sylvain Ribault (discusscontribs) 22:12, 27 November 2017 (UTC)

Ah yes, I disabled TOCs when first making the {{Article_info}} template. For the majority of WikiJMed articles it ended up looking odd. However, it's definitely a useful feature to have somewhere, so I'll add a parameter to reactivate it for longer articles. T.Shafee(Evo﹠Evo)talk 00:07, 28 November 2017 (UTC)
Now implemented by writing anything in the |toc= parameter of the {{article_info}} template at the top of the page. The expansion limit can be set by using a number (e.g. toc=1 will only show top level headings). I suggest a limit of 1 as default in order to keep them of manageable length. T.Shafee(Evo﹠Evo)talk 10:55, 28 November 2017 (UTC)
On second thoughts, its actually the minority of articles for which a TOC is inappropriate (e.g. this one). I've therefore made the default to show a TOC, and it can be hidden by setting |toc=off. T.Shafee(Evo﹠Evo)talk 11:51, 28 November 2017 (UTC)

That's great progress, thanks! But I still do not see why we show only the highest-level subdivisions rather than the full TOC. The full TOC would be very helpful for editors and peer reviewers, and maybe ordinary readers too. If the full TOC takes too much space relative to the rest of the article, the solution is maybe to restructure the article, not to hide the TOC. Well-structured articles tend to have beautiful TOCs of reasonable lengths. Sylvain Ribault (discusscontribs) 22:14, 28 November 2017 (UTC)

You're right that just 1 TOC level was was too short. I've put it up to two levels as a default, since three was unweildy (in the lead draft as my example). T.Shafee(Evo﹠Evo)talk 12:18, 29 November 2017 (UTC)
OK, I can survive with two levels. In the lead draft there is only one level at the moment, two would be better imho, but I am not changing this since I am not in charge. Sylvain Ribault (discusscontribs) 20:18, 29 November 2017 (UTC)

And why don't we display section numbers? We are used to having these numbers in Wikipedia, is there a reason for removing them? Section numbers can be very helpful when doing peer review. Sylvain Ribault (discusscontribs) 20:20, 29 November 2017 (UTC)

For numbering, I referred to a few other journals to check common practice. TOC numbers are only included if each section is also numbered. However, you're right that they could be useful for reviewers, so we could omit them in the published version, but include in the version as under review. T.Shafee(Evo﹠Evo)talk 23:23, 30 November 2017 (UTC)
Given that WikiJSci covers so many fields, I fear that the practices of existing journals may differ a lot. In physics most (all?) journals number sections, cf JHEP, JCAP, Nuclear Physics, SciPost, Physical Review. If journals do not have a common practice, why not follow Wikipedia? Sylvain Ribault (discusscontribs) 08:50, 1 December 2017 (UTC)
Browsing my own math articles I observe that sections are always numbered; TOC is present sometimes; "introduction" is numbered sometimes; "references" is never numbered. Boris Tsirelson (discusscontribs) 09:47, 1 December 2017 (UTC)
Really useful feedback, thanks. As you guessed, I've far less exposure to mathematical and physical science journals! It's really useful to have people involved who can bring different perspectives. So currently the TOC should be numbered and, on published articles, also stick to the top of the page when scrolling. In order to maintain VisualEditor compatibility, articles in review can't have the scroll effect. T.Shafee(Evo﹠Evo)talk 10:44, 1 December 2017 (UTC)

Academic standards vs Wikipedia standards[edit]

The quality standards for academic articles and Wikipedia articles (including Featured articles) certainly overlap: in both cases, the articles should be clearly written, in good English, and concise enough. There however seem to be significant differences:

Wikipedia Academic
Comprehensive Not necessarily
Neutral Not necessarily
No original content Possible original content
Preferably cite books, reviews Preferably cite original research
Cite sources whenever possible No references for well-known statements
Not necessarily Up to date, aware of latest research

There are good reasons for such differences. In particular, Wikipedia articles are supposed to permanently survive and evolve, while academic articles are practically immutable and destined to be superseded by later works. For an example of an encyclopedia article that is quite good according to academic standards, and poor according to Wikipedia standards, see the Scholarpedia article on gauge theories.

This raises the question: which standards should WikiJSci adopt? It seems inevitable that academic authors and reviewers will follow the academic standards they are used to, if WikiJSci is to attract people who are not already Wikipedians. If we wanted WikiJSci articles to follow FA standards in addition to academic standards, we would need an extra mechanism for that.

I would argue that WikiJSci need not impose Wikipedia standards in addition to academic standards, for the following reasons:

  1. The extra work and unfamiliar requirements would dissuade potential authors.
  2. The responsibility for enforcing Wikipedia standards is not WikiJSci’s.
  3. WikiJSci articles might have various uses in Wikipedia: They could be copied in totality or in part, or paraphrased, or summarized, or serve as authoritative references. If copied, they could be Featured Articles in some cases, and article of lower (Wikipedia) standards in other cases, to be improved later.

Sylvain Ribault (discusscontribs) 22:20, 28 November 2017 (UTC)

I would argue that you are contrasting original research articles and Wikipedia, while at WikiJSci the plan is to solicit both original research articles and review articles. My interest, admittedly, is more in review articles, since those are of most direct benefit to Wikipedia.
For review articles, the comparison is probably more like:
Wikipedia Review article
Comprehensive Reasonably comprehensive
Neutral Should be fairly neutral
No original content Existing content more important, some original content permissible
Preferably cite books, reviews Cite original research, reviews, books if applicable
Cite sources whenever possible One function is to give an overview of the literature, so sources crucial
Not necessarily Up to date, aware of latest research
So at least for review articles, I believe we should have harmonized criteria taking into account both the common scientific standards for review articles and WP standards. As the table shows, those are none too different to begin with.
Markus Pössel (discusscontribs) 09:53, 29 November 2017 (UTC)
I'll try to reply with more details later, however here are my general thoughts. I think that fully Wikipedia-integratable articles are some of the most valuable. These would have to conform to both Wikipedia's guidelines and the scholarly standards necessary for the journal's SCOPUS and PubMed applications. In particular:
  • Encyclopedic review articles in their entirety (either submitted from Wikipedia (example), or written fresh (example)
  • Focussed or figure reviews to have parts that are integratable (example)
However, not every article needs to be integratable into Wikipedia. Although such articles would miss out on some of the moat unique and powerful aspects of the WikiJournal format, they still benefit from the open access, low cost, collaborative nature, and can even be integrated into other Wikimedia projects. These need not comply with any Wikipedia policies, only scholarly journal standards. For example:
  • Original research articles that are never directly integrated into any Wikimedia project, but may serve as a reliable source (once the journal is established and appropriately indexed) (example)
  • Original research articles that can be used as teaching aids in Wikiversity (example)
There are also plenty of possible formats that we've not tried yet.
  • Plain language article that are suitable for those who do not have English as a first language could be integrated into Simple English Wikipedia (example)
  • Textbook-style walkthroughs to accompany a highly-technical Wikipedia article
That being said, we should absolutely try to make our guidelines as clear as possible for both editors and prospective authors. It's particularly important for authors to know what level of integration to expect for a given set of style guidelines. In general, I am willing to leave a lot of decision-making up the editors involved in processing individual articles (since even different areas of Wikipedia have different standards for e.g. number of references, and whether those should be primary, secondary or tertiary). T.Shafee(Evo﹠Evo)talk 12:15, 29 November 2017 (UTC)
It would indeed be good to have a lot of flexibility, given how widely practices differ across fields, both in the academic literature and in Wikipedia.
Now I am not sure that we really control the level of integration of a WikiJSci article into Wikipedia. We can certainly integrate an article into Wikipedia, but if we do not integrate we cannot prevent someone else from doing it. Wikipedia has a rule that says "If a rule prevents you from improving or maintaining Wikipedia, ignore it". In other words, an encyclopedic review article could be integratable as soon as it is better than an existing (or non-existant) Wikipedia article. It could happen that an article is rejected by WikiJSci and nevertheless integrated into Wikipedia.
By the way, what do we do with rejected articles? Erasing them (and the corresponding work) sounds dumb, I guess they will survive as preprints? Sylvain Ribault (discusscontribs) 22:32, 29 November 2017 (UTC)
They should certainly be categorised and labelled. I'll draft a template that can be added. I think authors should have an easy option to delete the rejected page, since it seems harsh to force retention of something that the author may no longer stand by. T.Shafee(Evo﹠Evo)talk 00:07, 2 December 2017 (UTC)
I've made a brief template here: {{WikiJPre_declined}}. It can be added to the top of any page, and will also be added by adding a |declined="any text" parameter in the main {{Article_info}} template. It also puts articles into the new category Category:Article_preprints_declined_for_publication_by_a_WikiJournal. T.Shafee(Evo﹠Evo)talk 01:03, 3 December 2017 (UTC)
On Arxiv, preprints are never deleted: on submission, authors grant Arxiv a perpetual license to distribute the text. In contrast to Arxiv, we expect preprints to receive comments. This is an additional reason for not deleting preprints, as we would have to delete the corresponding discussions too. On the other hand, on Arxiv, authors have the option of making drastic changes, and even emptying the text. But then older versions remain accessible, although not prominently. See for example this withdrawn preprint. Sylvain Ribault (discusscontribs) 13:08, 3 December 2017 (UTC)

Why allow confidential submissions?[edit]

In the pages on Submission and on Peer reviewers, it is explained that WikiJSci accepts confidential submissions by email.

A stated motivation for allowing confidential submissions is that some journals will not accept previously disclosed material. In physics and mathematics this motivation is non-existant: most articles start their life on Arxiv. Biologists too seem to warm up to preprints recently. And at WikiJSci, do we expect any submissions to be rejected and later submitted to other journals?

We could simplify the workflow, and the explanations to authors and reviewers, by not having this possibility. Sylvain Ribault (discusscontribs) 20:56, 1 December 2017 (UTC)

I know what you mean. There are fewer and fewer holdouts that don't allow preprints. For WikiJMed, the main problem was the New England Journal of Medicine. For WikiJSci, I actually can't find any big publishers that still don't allow submissions from pre-prints! However, I think it's a big enough issue that we should probably put it to a vote on the editorial board in case others know of something we don't. T.Shafee(Evo﹠Evo)talk 00:04, 2 December 2017 (UTC)

Another simplification that would follow from having all submissions public: as written in this reviewer invitation template, "Peer reviews for accepted articles are public". If all submissions were public, all peer reviews could be public too. Sylvain Ribault (discusscontribs) 21:51, 3 December 2017 (UTC)

@Sylvain Ribault:, your proposition is indeed noble and perhaps we do not have definite journals which do not accept publication material that have otherwise been previously open, we cannot ignore the possibility that there can be some notable or non-notable journals where the author can decide to re-submit. It would therefore be worthwhile to maintain the provision of submission by email. Nevertheless, we need not encourage it. Diptanshu💬 08:16, 8 December 2017 (UTC)
I am not proposing this as a matter of principle. Rather, I think that the journal's workflow, and the guidelines for authors and reviewers, should be as simple as possible, with no 'ifs' and 'buts'. Allowing special cases might lead to a net loss in submissions, because some people will be turned away by the apparent complexity of the process.
A possibility would be to display 'standard guidelines' on the main Authors page, with a link to another page for all the special cases, including private submissions. Sylvain Ribault (discusscontribs) 09:14, 8 December 2017 (UTC)

Using WikiJournal of Science as a reference in Wikipedia[edit]

Discussions have been held previously about WikiJournal of Medicine as a reference in Wikipedia ([1]), but W:WP:Identifying reliable sources (science) is not the same as W:WP:Identifying reliable sources (medicine), and previous discussions were held before 2+ reviewers became mandatory.

I've therefore started an entry at the "Reliable sources noticeboard" whether articles in WIkiJournal of Science can possibly be used as sources in Wikipedia:

Mikael Häggström (discusscontribs) 16:56, 3 December 2017 (UTC)

Editorial board vs Advisory board[edit]

The editorial board of WikiJournal of Science already has 16 members (including advisers) and there are still a few more unprocessed applications. Possibly we should decide upon whether we wish to set a soft limit on the number of board members we are willing to have. Furthermore, I think that we can have an independent Advisory board for the journal as well as for the WikiJournal as a whole. There are a few national or international figures whose experience could benefit us and yet we might not or they might not be very keen on having them/joining the editorial board. They can surely be accommodated there. The editorial board can nominate people who can be invited. Perhaps others too can place nominations while the editorial board finally approves it. The proposal needs further discussion and is not complete in itself. Diptanshu💬 08:25, 8 December 2017 (UTC)

I would invite the responses and inputs to be posted in the thread Talk:WikiJournal User Group#Editorial board vs Advisory board. -Diptanshu💬 16:35, 9 December 2017 (UTC)

Location of mailing lists[edit]

I have initiated a discussion Talk:WikiJournal User Group#Location of mailing lists, the scope of which extends to WJS. I would encourage the participants to join this discussion (please contribute directly to the mentioned thread). Diptanshu💬 15:31, 9 December 2017 (UTC)