Why restrict only for Scopus?

It is not relevant to use a paid database to measure open science (or in this case open access articles), 1) Scopus only index journals based on registration and include the ones that fit with their criteria, many journals from global south (like Indonesia where all journals are OA) are not included, 2) Scopus dbase contains mostly English-based journals. Again in this case, very few journals from Global South are listed, 3) Scopus does not index repositories. 1/2

Scientifc community can always go to Google Scholar, Base, Share database (and now Dimension) to get articles that are not listed in Scopus dbase. Promoting the usage of paid dbase to this project can lead to a more damage to publishing system in non US/EU countries, because they tend to follow the policy from western countries. 2/2

I would agree with Dasapta's comments here. By focusing on a single indexing & abstracting tool Open Science Monitor is immediately excluding a large corpus of relevant data and content. No indexing & abstracting service indexes all known literature so a multple system approach is a necessity.

Such approach may require the inclusion of DS Dimensions, 1findr (both of which are open) and WoS. Appreciate Google Scholar and MS Academic might be difficult to gather meaningful data from - but worth investigating? Whatever the final approach, something more holistic in scope would benefit OS Monitor, researchers and other users.

I agree with the other comments that paid/subscription databases should not be used in this analysis (or be the sole source). This is a massive opportunity to espouse the very open science practices that funders and institutions are aiming to encourage or mandate researchers to adopt.

The source databases should be (1) open, i.e. not behind a paywall, and with a clear, accessible methodology; (2) independent, i.e. not driven by a group that would benefit from it, but done in the public's interest, which probably means maintained by a consortium including universities; (3) multiple, as cross validation is key in capturing reliable trends. As such, Scopus has to be excluded.

As an open science advocate, I will not have a elsevier credential, I just erased my mendeley account for that reason. So if it is in Scopus, I would tend to think it is not open science. BTW: a login with orcid instead of twitter would have been nice...

To echo the comments above: closed databases are bad for open science. Please do not rely on closed databases (e.g. Scopus). Thanks.

Using a paid database seems problematic for promoting open communication within the scientific community... The other comments have provided reasons for the above claim.

I do not feel that making "# of Scopus publications" the basis for assessing the corpus of scholarly publications will result in a fair assessment of % of Open Access. Scopus requires that a journal publish for at least 3 years and then goes through a rigorous process where one of the criteria is whether inclusion of a journal will increase or reduce the reputation of the Scopus product. VERY many open access journals supported by university libraries are not indexed in Scopus. Many open access publications are new or experimental and would never be indexed in the Scopus database.

In addition to all of the above, it should be questioned why Web of Science is not being used for this process, especially now that it provides OA status after integrating with Unpaywall. WoS was also used for the first version of the monitor, so it is unclear why this switch was made, instead of using both services.

I agree that the choice of Elsevier as a subcontractor is very worrying, since their entire business model is either non-OpenSource, or very-expensive-OpenSource, both of which are counterposed to the openness of science, and draw important resources away from good research. It would be more credible to use a more independent subcontractor, which then can choose a more balanced set of resources.

get rid of Elsevier ! There data are fake : - "it is critical to understand that the Journal Impact Factor has a number of well-documented deficiencies as a tool for research assessment" SanFrancisco Declaration - ‘Predatory’ open access: a longitudinal study of article volumes and market characteristics Cenyu Shen and Bo-Christer Björk BMC Medicine201513:230 DOI: 10.1186/s12916-015-0469-2 - https://openarchiv.hypotheses.... - https://openarchiv.hypotheses....

Great initiative of the EU to establish an open science monitor. However, in its current form their are some troublesome limitations: 1) # Scopus publications: This point is not clear, as in brackets also other sources are given. I sincerely hope that the EU is not considering only scopus listed publications for the monitor, as several open science journals are not listed in Scopus. For example: "Journal of Open Source Software" and "Journal of Open Research Software".

2) Transparency and Open Access are among the most important benefits of Open Science. In contrast, the Scopus database can only be accessed with Elsevier credentials, making it impossible for Citizens Scientist without institutional access to check and validate any results from the monitor. In the light of current events, this might also apply to German and Swedish scientist in general!

Being a paid database and using non-inclusive criteria are not the only concerns here. Using Scopus as the primary basis for assessment will lead to disciplinary inequalities as well. The low coverage of Arts and Humanities journals in Scopus had been pointed out several times in previous studies. For instance, Mongeon and Paul-Hus (2016) (DOI: 10.1007/s11192-015-1765-5) concludes that the Scopus coverage of Arts and Humanities journals (compared to Ulrich’s extensive periodical directory) is less than 20%.

Also, in Arts and Humanities disciplines article citations are much less dominant indicators of excellence than in STEM. Therefore the inclusion of DOAB to the sources would contribute to a bit more balanced picture.

Assessing life sciences with Scopus and WoS would be acceptable. But social, educational, law, computational and economic sciences are poorly indexed by these databases. Indeed, history, educational, social sciences produce mainly book chapter, books and reviews, type of document poorly indexed by Scopus & WoS. Computational sciences produce mainly conference proceedings that are also poorly indexed by those databases. Relying heavily on local context such as history or human geography are not published in English international journals, and not indexed by Scopus & WoS. Therefore, source

2 Therefore sources for OA monitoring should be selected by domain. One size fits all is scientifically irrelevant.

Using the proprietary and for profit Scopus to monitor open science is like asking a fox to guard the henhouse in order to check the freedom of the chickens.

Scopus not the right source: a) to be included you need to have at least 3 issues, so for yearly journals it means 3 years b) monographs are never included into Scopus, and the coverage of Social Science and Humanites is not so wide. Using just Scopus means cutting off all half of the scientific world. At least DOAB (Directory of Open Access Books) need to be used, alongside with at least DOAJ (directory of Open Access Journals) c) Scopus is a proprietary database, so it can't be the only source. Morevoer, it's owned by Elsevier, which has been lobbying for years against Open Access

Unacceptable restriction to Scopus. At the very least it should be complemented by Clarivate Analytics data.

If this indicator is meant to measure all the "publications" in any discipline and language and not only "journal articles", the use of Scopus is highly misleading. Monographs are not included in Scopus nor are the vast majority of journals published in the Social Sciences and Humanities. Scopus is English-speaking oriented, thus excluding publications in other languages. The use of Scopus cuts off half of the disciplines in the ERA. Moreover, Scopus is a proprietary database, so it should never been used as the only source.

If we want Open Access results then we must ensure that the infrastructure underpinning them is built on the same principles. Placing the monitoring of Open Access output exclusively in the hands of a closed, for-profit service that ignores large swathes of academic output (e.g. non-English, AHHS-fields) not only undercuts the monitor's results; it will damage the credibility of the EU open policy goals themselves.

We all are aware of the importance (and power) of Scopus and WoS, but when will the EU open itself really to alternative indicators?

It is stated in the draft methodolological note that "the study covers all research disciplines, and aims to identify the differences in open science adoption and dynamics between diverse disciplines" (p. 6) I think it is hard to find someone today argumenting that one data source can accomplish this, a variety of sources is needed (and preferably open as stated in comments above). To pick one will take us back to the problem of searching the key where the light is, and this is not in tune with recent developments in the scientometric indicators field such as the Leiden manifesto (principle 6)

Why are all the sources listed for indicator P when, as I understand it, it will be based on Scopus solely? The same goes for the indicators TCS, FWCI, TP1/TP10.

I agree with many of these comments above (and I do not appreciate to have to login to make a comment, on top of that either with Facebook or Twitter). It is not good to base this only on Scopus (Elsevier).

Two main problems with Scopus as baseline for total pubs: it is not transparent and its coverage it not sufficient in terms of languages, document types and fields