Another small problem is that the data in the graph reflects the current situation for the various publication years, not the situation in the various publication years. That matters, because esp green and bronze open access tend to rise retrospectively over the years, because of embargos and publisher policy respectively. Stated differently: the current 72% for 2020 could perhaps rise to e.g. 75% for that same 2020 publication year over the next few years

There is a small problem here: the 100% ambition of the Dutch government relates to publications coming out of publicly funded research. The data used here looks at all Dutch research, including that which is not publicly funded.

Als er al iets duurder is geworden door Plan S is dat niet toegang tot de wetenschap maar het open access publiceren.

Open access staat niet gelijk aan betalen om te publiceren. Bijna de helft van artikelen in open access tijdschrift verschijnt in tijdschriften zonder article processing charges (de zogenaamde diamond open access journals). Ook het delen van artikelen via repositories is gratis. Dat laatste kan bij enige duizenden tijdschrift ook direct, zonder embargoperiode.

In de geest van open science zou je idd kunnen zeggen dat wat open kan idealiter ook open is, maar er is geen sprake van een vereiste van verregaande openheid. Zo zijn er in NL geen universiteiten die van hun onderzoekers hard vereisen dat zij open access publiceren. Enkele instellingen hebben wel een OA mandaat/beleid, maar zonder sancties.

Tot op heden heeft open science in Nederland maar heel weinig met onderwijs te maken.

Die zorgen zijn er inderdaad maar die hebben weinig te maken met open science en meer met uit handen geven van infrastructuur aan commerciële partijen en met gebruik van commerciële onderwijsmethoden

Dat doet te weinig recht aan de gigantische hoeveelheid rapporten, guidelines, framworks en principles die er al liggen en waar zeer veel denkwerk in zit. Veel open science practices werken al, dag in dag uit. Het is niet iets nieuws waar we pas mee verder kunnen zodra er weer nieuwe voorwaarden zijn bedacht. Uiteraard kun je wel terwijl de trein rijdt verbeteringen aanbrengen. En dat is precies wat er gebeurt.

Dat is een beetje een open deur. Zijn er belangrijke voorbeelden van waar open science productie en circulatie van kennis blokkeert?

Beperkingen op delen van data geschiedt vooral generiek: dus waar er privacy issues zijn of veiligheidsrisico's worden data niet openlijk gedeeld. Gesloten data kan uiteraard onder bepaalde voorwaarden nog wel weer 1-op-1 met bepaalde partners worden gedeeld waar dat kan binnen de privacywetgeving en veiligheidsnormen. Data die open wordt gedeeld is in principe open beschikbaar voor iedereen, wereldwijd.

Cornell betaalde in 2019 nauwelijks meer dan 10%. Vele instellingen betalen member fees en ook de Simons Foundation betaalt mee.

arXiv bestaat al 30 jaar, sinds 1991. Dat is dus niet iets dat nu ontstaat.

Deze "Guiding Principles on Management of Research Information and Data" hebben niet veel te maken met de FAIR uitgangspunten voor onderzoeksdata. De eerste gaan vooral over eigendom en beheer van data, terwijl de FAIR principes voor hergebruik en verificatie willen faciliteren. Ook gaat FAIR vooral over onderzoeksdata, terwijl de guiding principles zich richten op mogelijk samenwerkingen met commerciële partijen op het gebied van onderzoeksinformatie (dus informatie over onderzoek, niet informatie uit onderzoek).

Het zou heel goed die kan op kunnen gaan, maar dit is nu nog niet de werkelijkheid. Waar er wel al concreet iets gebeuert gaat het om metadata over onderzoeksoutput (zoals publicatiegegevens in hosted versie van CRIS systemen als PURE) en niet om onderzoeksdata. Maar het is zeker iets waar instellingen zich terecht zorgen over maken.

Ik neem aan dat wordt bedoeld: "verdienden voorheen" want zeker in Nederland is dat model voor de grootste uitgevers de afgelopen jaren al gewijzigd.

Er is voor zover ik weet ook niemand die dat ooit geoppeerd heeft.

Hier wordt neem ik aan bedoeld dat dit traditioneel gebruikelijk was. Huidige evaluatieprotocollen van zowel VSNU als NWO sluiten dit echter uit. Dat nieuwe systeem van erkennen en waarderen is daar dus reeds geïmplementeerd. Het oude systeem is informeel en voor status binnen wetenschappelijke gemeenschappen inderdaad waarschijnlijk nog wel invloedrijk.

Hergebruik van data is niet beperkt tot wetenschappers. Iedereen die dat wil kan de data bekijken en gebruiken. Afhankelijk van de gekozen licentie kan dat ook voor commerciële doeleinden

Dat moet (dus) 2020 zijn. De cijfers daarover worden in de loop van het jaar door de VSNU aan de minister gestuurd. Er is officieel nog geen beleid voor na 2020.

Op welke data is dat gebaseerd? Het klopt wel, maar afhankelijk van de gekozen database en jaar zitten het VK en Finland soms even hoog.

Nog voor het maken van de onderzoeksopzet zijn er activiteiten die ook relevant zijn in het kadfer van open science, m.n. de verkenning van onderzoeksprioriteiten. Het verkennen van wat onderzocht kan/moet worden is iets dat in toenemende mate deels open gedaan wordt.

Sinds 2016 is de horizon van het OA beleid is/was 2020, niet 2024. Daarbij laat de minister zich niet informeren over OA voortgang voor alle nieuwe wetenschapelijke publicaties, maar louter voor de peer reviewwed artikelen.

Welke keuzes zijn gemaakt bij die berekening? Gaat het om citaties in 2020 van artikelen uit alle jaren, of alle citaties van artikelen gepubliceerd in 2020? Kijken naar citaties van artikelen in 2020 op dit moment is een beetje hachelijk vanwege de korte tijd die verlopen is na publicatie. De kans om nu reeds geciteerd te zijn is simpelweg heel klein.

Het is lastig reconstrueren wat Science Guide hier precies gedaan heeft. Het lijkt er echter sterk op dat men ten eerste niet heeft ingeperkt op artikelen, terwijl men zegt wel een uitsrpaak over artikelen te doen. Ten tweede lijkt het er sterk op dat men het totaal aantal citaties ontvangen door OA publicaties uit alle jaren heeft gedeeld door het aantal OA publicaties gepubliceerd in 2020, en ingelijk voor gesloten publicaties. Dat zijn echter geen zinnige getallen.

E zijn vele (wetenschappleijke) onderzoeken naar het zogenaamde citatievoordeel van open access publiceren. Hoewel het zeer moelijk is het hard aan te toenen (want hetzelfde stuk is doorgaan niet zowel open access als niet-open access beschikbaar) geeft de bulk van de studies een positief citatieeffect van open access aan, voor artikelen en voor boeken.

The gold model has no advantage over the green model in terms of quality of research.

The green mandates are of course for research that would otherwise not be open because the gold route is not viable for a researcher. In almost all of those cases it is green that ensures funders that the research outcomes are available to all those who need them, as soon as possible, only hindered by embargo's.

Note that most green OA versions are after peer review and include any revisions coming out of that. Just as green versions, gold versions can never be fully guaranteed trustworthy, as for instance shown by the phenomenon of retractions.

There is absolutely nothing in the gold open access model that purposely aims at providing "full live and updated datasets". This reads like an enchantment of a publishers that wished datasets to become part of their ecosystem, while in reality researchers are absolutely free to publish thier datasets wherever they like. Making sure that they are attached to the scientific record is a matter of citing, preferably using sustainable data archives and perhaps of using permanent IDs. It has in itself nothing to do with journals.

All versions of all types of information are citable. It is up to individual researcher to choose what to cite. It information items are linked to related versions the reader can make up their mind about which versions to check and and build on. Green OA versions are routinely linked to (often closed, so to many people unreadable) versions on publisher websites. It would be good if publishers started to include links to green OA versions on their websites.

De vraag is of deze regeling de mogelijkheid biedt artikjelen in de repository libre open access te maken, met een open hergberuikslicentie (zoals CC-BY) die Plan S vereist. Zo'n licentie maakt in elk geval ook verkopen met een winstoogmerkt mogelijk en dat wordt hier juist uitgesloten

Interessant is dat van de artikelen+reviews uit 2017 in Web of Science met een Harvard auteur toch 'slechts' 62% open access is, waarvan 54% groen open access. Wellicht komt dit doordat de harvard license alleen wordt vereist indien de auteur een corresponding author is? Als dat zo is lost de license niet het probleem op dat hierboven wordt gesteld, nl. dat niet-corresponding authors geen gebruik kunnen maken van de lees-en-publiceer deals.

Zowel NWO als de EU (Horizon)/ERC maken het mogelijk APCs voor open access publicatie te betalen uit de projectbegroting.

In 2018 had echter zo'n 50% van de artikelen van Nederlandse instellngen ook een Nederlandse corresponding author, dus "het grootste deel" is hier misschien iets te sterk uitgedrukt. Ook is het heel goed mogelijkdat de buitenlandse corresponding author ook onder een read-and-publish deal valt of dat die buitenlandse corresponding author zelf uit andere fondsen de APC betaalt.

Wat wordt hier beoeld? Gaat dit om het zogenaamde bronze open access? Die vorm wordt beperkt toegepast, na een moving wall en zonder open licentie. Of bedoelt de auteur hier juist de gold open access, waarbij de uitgever een werk (vaak na betaling van een fee) direct en inderdaad vaak met open licentie beschikbaar maakt?

Dat is mij niet bekend en ik vraag me af of dat werkelijk in enige omvang is gebeurd.

Hoe succesvol? Is er een analyse bekend van hoeveel publicaties op grond hiervan in repositories beschikbaar zijn gesteld? Achterhalen van auteurs van pre-1997 werken en toestemming krijgen van hen is lastig.

De vraag is hoe het sindsdien is gestegen of gedaald. 2002 is 17 jaar geleden.

I do not think publishers are suspicious. Rather that readily introduce OA concepts where they find it does not threaten their profit levels and market control.

Not sure that is the case. It may rather be that funders take bold steps.

Not necessarily: they can also be hybrid journals if these are part of a transformative arrangement or in any journal if the accepted manuscript is shared immediately in a institutional repository, with a CC-BY license.

The distribution of APC levels is so skewed that the average is not very informative. Most big publisher APCs are 2500-3000 USD while those in journals from smaller society and institutional publishers are zero or very low.

That is not true. Over 70% of open access journals do not charge a publication fee.

A consistent readership level is not a criterion for listing in the journal citation reports and getting an IF.

AFAIK there is no evidence that this is the case. It might hold for the very small number of general science magazines (Nature, Science, etc) but not generally.

Not quite invariably, there are many young journals with higher IFs than older journals

Just for the records, many funders, institutions, societies and publishers agre that judging researchers and research by impact factor is not good practice. See https://sfdora.org/

Not sure that is the case: researchers want their paper online, but care less about the rest of the journal. Most are still quite content with hybrid journals, alas.

And many would add: available for reuse, e.g. in course packs, in collections ready made for text-mining, in slide decks that researchers share etc.. This requires an open license, like CC-BY

The suggestion that scientific integrity and open access are antithetical is unfounded. Open access and open access journals have been tested for decades. Worrying about quality and evidence based information is good and necessary. However, that requires more openness and transparency, not less.

Or because of the limited open availability of mainstream science in paywalled journals?

This is a whole separate and complex issue, having to do with registration of trials, publishing null and negative results, sharing data, code and workflows and having rigorous checks for soundness..Simply not publishing open access does not help at all, on the contrary, one might say that open access journals in medicine (e.g.PLoS medicine) are likely more advanced in requiring open science than paywalled journals.

Interesting point. New for me. Would like to see that tested. Are APC-based open access journals having more difficulty finding reviewers? I have seen boycotts of researchers not wanting to review for Elsevier, but haven't seen those for reviewing in APC based OA journals. But as said: would love to see data on this.

That is seriously questioned, as many irreproducible results have been published in papers in medicine, biology, psychology and economics despite allegedly strong peer review.

This is a fair critique, though only holds for (expensive) APC based publishing. Green OA and diamond journals do not involve paying APCs. Also, Plan S demands waivers for low income countries.Still not perfect, but together with the deceasing need for subscriptions this is better than what we have now. At the same time we (and the Plan S funders) need to much more promote diamond venues and the route of preprints with separate reviewing.

Why is the gates foundation the main inspiration? Maybe I missed something?

And that is very good indeed. As long as editors do not get bonuses for impact factor rises of course ;-). Also, at full OA journals run by legacy publishers and learned societies, would you say these incentives are leading to publishing substandard papers?

True, this is a real danger, but cOAlition S is aware of it and tackling the impact factor craze.We need to be vigilant of how this develops.

No, in a change from subscriptions to APC based publishing the costs can remain at the same level: same income for publisher, same costs for academia. One could and should go further and make the system way cheaper, , but basically there is no extra money needed.

Whatever way, academia is footing the bill. What one gets for the money remains the same, or actually, one get a lot more when paying for open access, namely open access, also for all those currently not in reach of subscription access. But yes, the costs at NEJM are stratospheric and should come down substantially.

Their higher staff remuneration is unneccessarily with the editor in chief making USD 700K: https://richardswsmith.wordpress.com/the-new-england-journal-of-medicine-open-access-plan-s-and-undeclared-conflicts-of-interest/

Another unsubstantiated claim.The bulk of the journals in the DOAJ do not charge APCs. So what incentives are we talking about there?

The claim get harsher here: provide evidence, or at least numerous results or leave out this assertion

There is no evidence supporting this harsh claim. I would leave it out.

The Lancet itself claims it is compliant (though actually it's the authors self archiving their OA paper): https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(19)31322-4/fulltext31322-4/fulltext)

Science publsher has been suggesting they might allow immediate green OA as a comliant route: https://www.sciencemag.org/news/2019/05/meet-plan-s-open-access-mandate-journals-mull-setting-papers-free-publication

No it does not, there are several options, together covering nearly all journals with a viable and compliant open access route https://101innovations.wordpress.com/2019/06/15/nine-routes-towards-plan-s-compliance-updated/

There's also an open letter, with currently 1928 signatories, that support funder open access mandates: http://michaeleisen.org/petition/

There is no principle anymore that banishes publishing in hybrid journals

No it doesn't. It says"must be published in Open Access Journals, on Open Access Platforms, or made immediately available through Open Access Repositories without embargo". See: https://www.coalition-s.org/principles-and-implementation/ The implementation also leaves room for conditional hybrid open access publishing.

10 months ago ....

Actually it is denominated cOAlition S. Science Europe coordinates the initiative but is a much larger group, see: https://www.coalition-s.org/funders/ and https://www.scienceeurope.org/about-us/members/

Plan S does not mandate publishing in open access journals

NIH does not mandate publishing in open access journals

Connecting in this way predatory and genuine open access is quite disingenuous

generally the non-predatory open access journals (the great majority) do not send you soliciting spam

not part of the implementation

rather 0-6000

In Plan S it is stipulated that researchers do not have to pay: that is for funders (or institutions).

The 4th Plan S principle already takes away this burden from cOAlition S funded researchers. What would exactly be the reason to endorse the FAIR OA principle additionally? Is it a responsibility for the science system as a whole? A potential danger is that the system of journals and venues important for researchers for the global North is being supported made fully accessible in terms of reading and publishing, while journals from lower income countries may be deemed non compliant while only charging an APC of say 100 Euro. These kind of unwanted effects must be carefully studied and avoided, next to the unintended consequence you already mention.

This indeed needs attention, but needs to be devised and agreed on by a much wider group of organisations than just cOAlition S. It should not be left up to one cOAlition of rich funders to decide on what is efficient and equitable. Also it would be good to have much more research into availability, use, effect and appreciation of the various types of waivers.

Although I also would recommend looking into this I would say that current "technical state of play" is in itself not sufficient reason to be more lenient. Of course there needs to be a more elaborate reasoning by the cOAlition of why these technical standards are necessary, who can best provide content with these additional characteristics and what is a reasonable time frame, given that publishers are given to option to wait with flipping until 2025 (if they sign transformative agreements).Perhaps there could be an analogous kind of transformative agreements for repositories as well?

Indeed there will be a growing role for third parties monitoring costs and other venue characteristics. DOAJ and Sherpa/Romeo are likely candidates but others can take a role as well. DOAJ has already announced changes to their governance structure. To keep things manageable the full cost transparency requirement could be limited to those venues with an APC above a certain threshold (say 700 Euro).

One consequence of not flipping is expected: those journals will not receive any more submissions from cOAlition S funded researchers from 1-1-2025, which by then may have grown considerably. The exception is that, depending on the outcome of the 2023 evaluation, papers may perhaps still be submitted and self-archived. Another development may be that of libraries starting to cancel subscriptions. That depends though on many other factors (OA share in the journal, prices, library budgets, willingness of libraries and their institutions to take risks)

I think this is exactly what the cOAlition S members would love to see happen more. The plan nowhere asks standards of curation, peer review, copyediting, and formatting to change. Reputations can be kept asd, transferred. Excellence is a concept that I think we should get rid of. It has been argued that the concept has no intrinsic meaning: http://doi.org/10.1057/palcomms.2016.105

No need to single out Norway. This is about 13 countries, with a number of them fully or almost fully: Norway, Finland, UK, France, Netherlands, Luxembourg, Austria

That is an interesting observation, as the organisations of young researchers have ushered their support or are critical of an APC-based system but want to go even further in changing the system: https://globalyoungacademy.net/young-academies-release-statement-in-response-to-plan-s-on-open-access-of-scientific-output/

That is not necessary, there are many solution and organisation in place that can help you out, e.g. https://www.fairopenaccess.org/ or a low cost publisher/host like https://www.ubiquitypress.com/site/publish/ and many more. Your library can probably guide you in this as well. I would urge you to consider this, perhaps together with editorials boards now at journals with high profit commercial publishers. That way we can make publishing cheaper, more open and fair.

This is not what is suggested or longed for

Yes that is a major issue. The current inequality of the subscription system denying researchers access must not be supplanted with another one making it hard or impossible to publish. That is why diamond routes (fully OA journals without APCs or preprints + overley PR) are to be preferred.

Nothing needs to change if the journals swap. New journal can acquire reputations and support communities as especially in the humanities has been the case continuously. More fundamentally: selecting papers to read or researchers to grant money to based on journal covers is a fundamentally flawed and harmful practice.

Please note (I hoped that was common knowledge by now) that OA journals are about curated peer reviewed articles. The juxtaposition of tradtional journals as high quality and peer reviewed and open access journals as low quality and not peer reviewed is not realistic.

This is indeed a challenge, but one that the publishers could have seen coming for many years. Ideally that knowledge and community does not get lost if either journals are swapped to full OA or if editorial boards leave the publisher and make an open but just as good journal. That has been done with success.

This is not the case. The principle are the same for all (with exceptions for books and transitionary deals), but leave a lot of room for varying solutions to be compliant: https://101innovations.wordpress.com/2018/10/22/eight-routes-towards-plan-s-compliance/

One other correction: you mention @OUPAcademic and @CambridgeUP as charities. However their 2017 real paid average hybrid APCs of €2437 and €2137 are on par with Elsevier's of €2615 and higher than Taylor&Francis/Informa's of €1796 acc. to @oa_intact https://treemaps.intact-project.org/apcdata/openapc/#publisher/period=2017&is_hybrid=TRUE

Is it? Have not read that. In is not in the principles. And the additional communication mainly targets companies sustaining the hybrid model without transitioning and companies with excessive profits. But it is not singling out named companies. The plan calls on all companies with these traits to change.

Why is that a problem? With all change (voting rights, climate control etc.), it cannot be universal from the start. The differences that creates during the transition can be overcome. Also, with these global transition, it is not unlikely that those countries that make the transition early will in the end profit from that.

A few corrections: the main Dutch funder @NWONieuws (including @ZonMw for the health/medical domain) is fully in cOAlition S. Two Swedish funders @forteforskning and @FormasForsk are also part of the cOAlition.

This does not answer the question. Referencing the data does equate to openly (with open license) sharing the raw data used in the monitor. The answer given does not guarantee the possibility to to independent inspection and analysis, while that would be easily possible. It consider this one of the main points that need to be stressed.Of course the consortium can still choose to share the data, protocols and scripts, but apparently the EU will not demand that .

Good to see that the commission implicitly admits that the monitor has limitations because of non-open nature of the data

This is important and offers and opportunity to see whether the choice of subcontractor and the chosen data sources have played a role or were discussed in the procurement.

This does not answer the question. If "collaborative" means that anyone can be a collaborator rather than just the consortium members and the subcontractor(s), those collaborators should have insight into how their contributions (e.g. comments on the indicators) will be weighed and dealt with.

We are not asking to make things better (though there are many opportunities there as well, like openly licensing citations citations) but to make things at lower cost, open, and reusable.

As an alternative, you could work to publish the book as a beautiful CC-BY Gitbook, free to access and reuse for anyone, thus contributing to lowering the cost of education. Especially for textbooks this is the way to go.

About security risks of open data: no one has ever advocated that data that poses serious threats when published should be made open. Relating this to open science is irrelevant, again bordering on malignant. Exceptions to opening data are evident and accepted.

The section on science communication has nothing to do with open science. Being precise, honest and communicative has always been important in communicating about (complexities of) research.

Regarding predatory journals: yes, there is a problem, though minor (way less than 1% of publications). Also scientists always needed to check the quality of journals, open access or not.

"The open access publishing movement, which demands payment from authors". That movement is not demanding payment. APC are seen as just one option to provide open access. They are often paid by institutions. Financially it should reduce overall costs.

Hornbostel syas "Storing data is time intensive and costly". Indeed it quite often is. But not being able to reuse the data, being unable to reproduce results or being unable to perform replications or meta-analyses is much more costly.

As far as I can see Hornbostel is misquoting the survey on (a.o.) peer review. The question (v5.8, p. 22) was not just for those with experience as peer reviewer. Also it was about quality of reviews, not reviewers. And the number (40%) should be 21% (p. 96).

Suggesting a link between experienced lowering of peer review quality and Open Science is nonsense. That survey was not on open peer review, but on peer review in general and thus mainly about traditional peer review. The suggestion is malignant.

Revealing reviewer identities: this is still very rare and to my knowledge there is no evidence to its supposed detrimental effect. Perhaps more importnat than open reviewer identities is to have (just) the reports open.

No. Before the movement towards open science the science system was not characterised by trransparency: papers were/are paywalled, peer review is invisible, funding proposals and evaluations fully behind closed doors.

The piece itself states that trust in science hasn't dropped, sort of disclaiming all the rest that it put forward.

Ranking and monitoring institutions has nothing to do with Open Science. If anything, open science advocates plea would be to move away from simple metrics, move away from hollow hierarchies of journals and institutions.

Increasing public trust is not the only and perhaps not the most important goal of open science. It is also about verifiability, reproducability, efficiency and relevance. It aims to do that by sharing data and code, publishing negative results and by public engagement

This article has many flaws. It is quite suggestive, which is aggravated by the choice of the title.

There is a very big difference between refusing commercial entitities a role in the open science landschape and giving the commercial publisher with the largest vested interest in closed science an undisclosed role in monitoring the progress of open science....

It is not clear what exactly Elsevier's role is. Who is setting goals, deciding on variables, on indicators, on data, on cleaning, on computation/analysis, on presentation, on outreach? The consortium should make clear exactly what is the role of Elsevier as subcontractor.

Avoiding bias in OS assessment is not a question of using one company's metrics for one variable and another company's data for another variable. It is not that simple!

These metrics may be useful to some but they have next to no value for open science.

Absolute size numbers are not a valuable argument when wanting to demonstrate your OS inclination. More interesting would be relatieve numbers. What percentage of your papers is open access with an open license, what percentage of journals have you flipped to full OA?

Absolute size numbers are not a valuable argument when wanting to demonstrate your OS inclination

Elsevier director of scholarly communications told this holds for all preprint archives, see this tweet

Not sure about that as the platform is only for Horizon2020 funded output and expect to have only some 10% of that. So in the full EU paper output it will be <1%. Nevertheless, if all funded adopted this model it could become substantial.

Dit misschien breder maken? Niet alleen impact factor, maar alle journal based metrics.

This may be overstating it. Yes, they need to be managed, people need to be aware of their existence, but once installed good and well implemented proxy tools only ask to login once during your browser session and only if you click paywalled content content. I suppose in that regard RA21 will be the same.

Dit is raar: in alle journals van de database of open access journals kan open access gepubliceerd worden. Duh.

Voor de uitwerking daarvan zie het nationaal plan open science: https://www.openscience.nl/binaries/content/assets/subsites-evenementen/open-science/nationaalplanopenscience_nl.pdf

Als de insteek van het verhaal fact checking is zou is deze uitspraak graag gecheckt zien. Betrouwbaarheid van peer review is lastig meetbaar. Peer review grotendeels een black box. Artikelen uit tijdschriften met een hoge impact factor zijn zeker niet zonder meer betrouwbaarder dan die uit tijdschriften met een lage. Er is zelf een psoitief verband gevonden tussen impact factor en het percentage retractions van artikelen.

The M in STEM stands for medicine, not for mathematics, which is considered part of science

This claim is unsubstantiated. The reference you give for this also makes the claim without providing evidence or data. The research by Richard van Noorden you refer to further down has finally corrected this false view, but you seem to have a reason to give greater prominence to the unsubstantiated claim. Why is that? Also the number you quote from Van Noorden's paper is 0% for all paper 1900-now, but the more important figure is that currently fewer than 10% of the papers will remain incited.

Actually you should read this as: "Yes but only by unaltered copying". Choosing this license means people can't rework/rearrange your content.

Actually you should read this as: "No (excepting private use)"

Trachten alle gouden open access publicaties te tellen zou ik niet willen kwalificeren als marchanderen met de doelstellingen. IN 2016 is de categorie hybrid (die Rathenau dus niet meetelde) even groot als de door Rathenau getelde pure gold artikelen. Met alle OA zitten Nederlandse universiteiten in dat jaar gemiddeld op 42% volgens Web of Science. De VSNU komt op basis van eigen bronnen en eigen tellingen ook op 42%.

Dat heeft Dekker intussen bijgesteld (samen met de EU ministers) naar 2020

Dit moet zijn: tweevijfde. Van de output van 2016 is 42% open access. Zie het bericht van de VSNU zelf: https://us7.campaign-archive.com/?e=&u=6e6efd9c9c87905323ae817ec&id=4553fa5fe0

Alternative title suggestion: We want Open Science to develop as slowly as possible

Indeed we should not use market power to buy up all potential competition

yes we indeed should avoid duplicating expensive, closed platforms and publication venues

In social science the longer format is not the default

Alternative title suggestion: "We want more and longer embargos!"

Alternative title suggestion: "We want more money!"

alternative title suggestion: "We want to stick to our business models!"

This is not logical. A n OA mandate combined with APC cap does not hinder overall growth of OA. It just means that soms journals have to change their APC levels to remain competitive.

that is strange indeed

I do not think the Ottawa Citizen has access to that.

Is there any evidence on the numbers?

That is changing. Many institutions and funders just ask you the mention what you consider your best papers/output

Zie hiervoor bijvoorbeeld het begin dit jaar door alle partijen onderschreven Nationaal Plan Open Science, paragraaf 3.3. Ook hebben alle instellingen via de VSNU DORA onderschreven waarmee zij beloven impact factors niet meer te gebruiken voor het beoordelen van artikelen of personen. Ik ben het er wel mee eens dat druk van de overheid nodig blijft om universiteiten en NWO/ZonMW hierbij op koers te houden.

Er zijn vrijwel geen subsidieverleners meer die expliciet om impact factors vragen, noch bij de EU, noch bij NWO. In praktijk zal het tussen wetenschappers best nog een rol spelen, maar het is geen beleid van de overheid of van de funders.

We zitten nu ergens tussen 30-40%. Bij 8000 tijdschriften kunnen auteurs van Nederlandse universiteiten zonder directe kosten OA publiceren. Dat zijn juist die tijdschriften met die impact factors. Of dat goed is en het geld waard is is waarover je kunt twisten, maar het is dus niet zo dat in Nederland nu het assessment systeem (voorzover dat nog IF-gebaseerd is) open access tegenhoudt.

Die oproep is al wat ouder. In 2013 maakte de regering bekend dat het in 2024 100% open access van Nederlandse publicaties wilde in 2024. Dat werd in 2016 versneld tot 2020.

That ties is nicely with the LIBER principles for negotiations: http://libereurope.eu/blog/2017/09/07/open-access-five-principles-negotiations-publishers/

At some point Wikimedia's Scholia initiative might be able to play a role: https://tools.wmflabs.org/scholia/

Even long before the second world war there were journals that individuals could subscribe to and that were available in academic libraries. Exchanging written letters as the dominant form of scholarly communication seems a bit older than before WWII. But I must admit that I do not have references at hand supporting that. Would be interesting to find out.

Can you name which ones did so? Would be very interested.

That is not correct I think. They just publish generally before journal organized peer review

Albeit not really at the global level yet I think the Vienna Principles and The Scholarly Commons with their principles do provide overarching goals.

Many thousands of journals are still published by institutions and societies operating as non-commercial entities.

Not entirely fair of course. The comparison should be with Google Scholar, which does limit itself to scholarly pubs (mostly), does indicate open versions and has a broader coverage and full text search compared to Open Knowledge Maps.I'm always in for new tools, but we need to make a full comparative review

We have put some of the definitions into context here: https://im2punt0.wordpress.com/2017/03/27/defining-open-science-definitions/

Looking forward to that and eager to learn how it will compare with the Dutch National Open Science plan. I hope it goes a few steps further. https://www.openscience.nl/en

Anyone can already provide that context by annotating and commenting using a platform independent commenting tool such as hypothesis https://web.hypothes.is/

Almost all ("standard") journals (including those from the big publishers) allow one to pay a fee to make an article open access.

The international Force11 community developed the pirnciples for such a commons and provides guidance on how it could function:https://scholarlycommons.org/

Een belangrijk punt. Is hier onderzoek naar gedaan?

Dat lijkt een zwak argument: als citaties/gebruik van boeken wel meegeteld zouden worden, was het probleem dan opgelost?

Dit geldt natuurlijk evenzeer voor natuur- en levenswetenschappers

Met het tekenen van DORA heeft de VSNU beoordeling op basis van standing van tijdschriften officieeel afgezworen en NWO vraagt er ook nauwelijks meer om bij subsidieaanvragen

Internationaal wordt hier ook over nagedacht in het project HumetricHSS

Wonder if you also include institutional and government open science policies that have been released or are being drafted

This/our global survey results on tool usage (as proxy for practices) may be useful: https://101innovations.wordpress.com/survey-2015-2016/

of course these are only examples but if these were the main or only ones that could introduce a bias ;-)

for a database of tools, amongst which many OS oerient tools see: http://bit.ly/innoscholcomm-list and you probably know about scicrunch

Are you aware of this report by Knowledge Exchange? http://www.knowledge-exchange.info/event/ke-approach-open-scholarship

In nieuwsbrief 2017-2 stelt Ton van Rietbergen dat ik zou propageren om met Wikipedia je H-index op te krikken. Dat is het laatste wat ik zou adviseren. Ja, draag bij aan Wikipedia maar nee, doe dat niet met als doel een H-index op te krikken. Het werkt niet en en moet niet.

Do you have indications that it is currently? That would be very interesting. What indicators would we use for that? Sorry for the avalanche of comments/questions ;-)

Which ones are you referring to? FAIR principles? Vienna Principles? others?

Exactly. Which is why we need to work on changing reward & incentives. And need to focus on showing how you can profit form OS done by others as a way of convincing people to apply OS practices themselves

I agree that there may be half baked requirements, but we need to make clear which ones we are talking about. Is it FAIR, ORCID, pregiststrations?

Are you referring to any particular research? Is that available/published?

I wonder what you mean by this. Is this to say that Open Science is irrelevant for those doing qualitative research? Or something else?

STM has >120 members. Will this mean 120 cases?

That could be goodwill, but it may also be out of fear that more researchers will get more accustomed to using Sci-Hub

As is providing information only after payment

Gold open access is not equivalent to fee based publication. Gold OA means OA provided by the publishers (as ooposed to green OA which is provided by the author). The are many GOLD OA journals, the majority in fact, that are not asking an APC.

I like that

Interesting. Does this mean that there are cases where research that may have taken for instance 4 years will remain unavailable for the research community for another 6 years? Let's try to make a list of pros and cons of that.

I wonder whether these or other journals use the graphical badges in their journals. As far as I can see, no journal does this, although many include contributorship information in textual form.

This call for Open Science is very welcome. It provokes scientists to think why they work the way they work and what holds them from being more open. I applaud trying to do this in a radical way accross your whole research cycle. That in itself may be an experiment and relatively rare. But taken separately each aspect of Open Science is not that rare. Radical full cycle openness may be at 1% but partial open science is probably already at 5-20% (just my guesses). So not mainstream, but certainly not marginal:

Publish your grant? Try RIO: http://riojournal.com/

Early sharing of your paper before journal publication? Look at the millions of preprints here: https://osf.io/preprints/

And that journals allow this has been known for years and can be checked here: http://www.sherpa.ac.uk/romeo/index.php

Publishing your data? Loook at the millions of data sets already shared in the 1000+ data archives here: http://www.re3data.org/

Sharing your brief ideas for research? Look at JoBI: http://beta.briefideas.org/

Opening up the notebooks of your experiments? Look here: http://onsnetwork.org or here: http://www.openwetware.org/wiki/Main_Page

Sharing video's of your experiments? Watch this: https://www.jove.com/ (This one is not open by the way)

Publish your workflows? Try this: https://www.myexperiment.org/home

Share your code? Well you know about GitHUb of course: https://github.com/

Online and open drafting? Many options again, try Overleaf, Scalr of Authorea

I could go on and on, for more than 600 tools look here in the list we created: http://bit.ly/innoscholcomm-list

Want to be able to read this as a comment on your NRC blog of 20150415? Install the universal comment layer as browser extension: https://hypothes.is/

Some of these tools have been around for a decade or longer and most are used by thousands of researchers worldwide. Join the club!

This title may be somewhat misleading as the VVD lost a lot of votes and the PVV won a lot of votes. It's just that the latter did not manage te become the biggest party. But the 'bad' populism as you call it is still on the rise in the Netherlands

It would be even more impactful to have all abstracts of literature indexed bij the BASE search engine translated into 5-10 major languages. And even further: to have a system attached to that for requesting translation of the full text. There must be a smart networks solution for arranging and funding that.

That solution mixes up 2 problems: jargon and language. They deserve seperate treatment.

Base search does provide some multilingual search support: https://www.base-search.net/?l=en

This reads like an implicit suggestion that non-English documents have lower quality. Is there any evidence for that?

This reads like an implicit suggestion that non-English documents have lower quality. Is there any evidence for that?

Of course they may also have chosen to publish in English because they want their research to have a wider audience

Better late than never, just a minor correction: the journals in which Dutch affiliated authors can publish OA without any cost for them in 2016 was 141 out of 2400 (not 5331), amounting to 5.9%. Now (from December 2016 and through 2017) it is 276 out of 2400, so 11,5%. And at the end of 2017 another 140 will be added. But sure, that still is of course much worse than what other publishers agreed on.

For script sharing based on these data also see: https://www.kaggle.com/bmkramer/101-innovations-research-tools-survey

I do not agree that these globally scoped repositories exacerbate the problem. By promoting and facilitating data to be open that were hitherto not available at all, at least they contribute to findability. Yes, they do not curate, moderate or integrate, but most of them do adhere to some standards at least (e.g. CC-0) and some of them do provide for and stimulate good metadata use.