All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Thank you for making the corrections and carefully addressing the (most helpful and attentive) referee's comments.
[# PeerJ Staff Note - this decision was reviewed and approved by Dezene Huber, a PeerJ Section Editor covering this Section #]
I only sent the paper to the more thorough referee. They are positive about the revised version but point out some possible errors or issues requiring checking or clarification. Please attend to them as appropriate.
I also suggest you rotate Fig 1 so the journal names are not all sideways and will thus be easier to read. Note the "Lorem ipsum" on Fig 2 which I assume is a mistake. Your results are very interesting and should cause reflection amongst scientists and taxpayer-funded institutions about the value they get from APC and Subscriptions.
Much improved, thank you.
I will use this text box to respond to some comments made by the authors in the response to the first round reviews:
* The authors state that: "It appears that no members of the faculty were removing or biasing their [Google Scholar] profiles in any way that would affect the results of this study." does not strike me as particularly rigorous.
Just so that the authors understand the completeness problem here, I will laboriously explain it:
When logged-in to your Google Scholar profile, for any publication currently showing on your profile as one of your publications, you can click on it, then click “edit”. After clicking edit it takes you to screen where options are given including “Remove this article and its [XX] citations”. Similarly, if one has a common surname, Google Scholar can regularly give you an option to ‘claim authorship’ on a publication and have it listed in your Google Scholar profile as if you had authored it (even if you haven not!).
Google Scholar profiles therefore do not necessarily reflect the true publication profile of an individual as publications can be hidden/removed from Google Scholar profiles on whim AND in some cases it is also possible to _add_ publications to Google Scholar profiles that were not in fact authored by the person.
There is obvious incentive for people to delete publications they might consider 'embarrassing' from their public Google Scholar profile, and Google Scholar makes it very easy to do this. Asking a person face-to-face if they have ever selectively removed a publication from their Google Scholar profile is not an effective way of determining the truth (people lie!).
A spot check using an alternative independent (non-Google) data source on one or two scholars, using their name string and/or their ORCID might be a good way of reassuring readers that the Google Scholar profiles analyzed here are indeed a complete record of journal publishing events by those faculty, but this is merely a suggestion and I won't insist on it - it probably doesn't affect the results of this study, but would help to give readers the impression of rigor and awareness of the trust/completeness issues with Google Scholar profiles.
I thank the authors for a much improved manuscript and dataset.
Clarification: the "Average citations per year" column in the collected data.
Does this:
a) capture of the average citations (according to Google Scholar) for each and every paper in the journal
or
b) capture the average citations (according to Google Scholar) of _only_ Kansas EEB authored papers in the journal
I think it would be good to better document and clarify this unusual and non-standard measure.
``````
I think it is mis-framing to talk write in this manuscript about the 'cost of openness'. I think it more accurate to discuss the pricing of routes by which open access can be achieved. APCs are prices not costs. The publisher chooses a price to set, a price which may be entirely divorced from the basic costs of doing the work to the publisher.
Line 204: “Finally, only for the purpose of illustration, we obtained what we term a “cost of openness” for each journal. This quantity is the APC in the case of open access journals, but is the hybrid open access fee in the case of subscription journals.”
Although this route is not available at ‘pure’ open access journals, I must feel I should point out that most/all of the “hybrid” (paywalled) journals permit the green open access route (self-archiving). There is a cost of $0 associated with the green open access route. The analysis presented does not seem to factor-in the availability of green open access routes.
One of the authors listed on this manuscript (Josh Bolick) definitely knows the in’s and out’s of “green” open access and has published excellent work on it e.g. Bolick, J. (2017) Exploiting Elsevier’s Creative Commons License Requirement to Subvert Embargo. Kraemer Copyright Conference Poster Presentation. Colorado Springs, CO. June 5-6, 2017. I’m thus at a loss as to why green OA is so ignored in this manuscript until a very late and inconsequential mention at the end. If the central thesis is that hybrid APC pricing is appallingly high (and it is!) -- would it not be helpful to point out to colleagues that open access can also be achieved at a cost of $0 via green open access, and particularly via perhaps under-utilized methods such as the Bolick manoeuvre (Bolick, 2017).
I'm still a bit concerned about the validity of some individual statements in this manuscript and the data.
a) the seven journals named that according to this manuscript have charged Kansas EEB faculty over $9000 to make a single article open access within them.
I know that the list price for hybrid OA at Nature Ecology & Evolution, and Nature Plants is over $9000 - those I do not blink at. But for the other 5 journals named, the list prices given at the publisher site don't make it at all obvious how Kansas EEB faculty were charged over $9000 and so I think they need checking and or an explanation given (even if it is just an explanatory note in supplementary materials).
For Aquatic Insects (Taylor & Francis), if the paying authors are located in the United States, the list price for hybrid OA for all articles types (“research articles” “original paper” “method” and “book review”) is at this point in time $3500 https://authorservices.taylorandfrancis.com/choose-open/publishing-open-access/open-access-cost-finder/?category=all&journal=naqi&fulloa=0&openselect=1¬available=0&dove=1&routledge=1&tandf=1&numberofresultsperpage=5&pagenumber=1
For Annales de la Societe Entomologique de France (Taylor & Francis), for hybrid OA, if the paying authors are located in the United States, the list price for hybrid OA for all articles types is at this point in time $3500 https://authorservices.taylorandfrancis.com/choose-open/publishing-open-access/open-access-cost-finder/?category=all&journal=tase&fulloa=0&openselect=1¬available=0&dove=1&routledge=1&tandf=1&numberofresultsperpage=5&pagenumber=1
Aquatic Microbial Ecology is published by Inter-Research. For this journal the hybrid APC pricing information is clearly given on their website: https://www.int-res.com/journals/ame/about-the-journal/#tab2box
“Articles submitted on or after 1 November 2022: €2500 for Research or Review articles
Articles submitted on or after 1 June 2021 and before 1 November 2022: €2000 for Research or Review articles”
How did Kansas EEB end up paying over $9000 for open access for a single article in the journal Aquatic Microbial Ecology? I cross-checked with the openAPC database which has records of over 190,000 APC payments contributed from over 400 research institutions. No university has ever paid anything like that to Inter-Research for a single article, according to openAPC data. https://github.com/OpenAPC/openapc-de
Deutsche Entomologische Zeitschrift is published by Pensoft. Their pricing information is clearly explained here: https://dez.pensoft.net/about#Core-Charges
For an article that is 41-50 pages long the APC pricing is listed as € 1,400 with a charge of €15 euros for each additional page above that. By my calculations one would have to submit a manuscript with over 472 pages above the 50 page limit, in order to rack up a total bill in excess of $9000 for a single article in this journal. Whilst this is within the realms of possibility, it seems unlikely and perhaps deserves explaining to readers the exceptional case of a very lengthy article if that is indeed the cause of the high bill in this case.
These are very specific assertions about businesses and if they are factually incorrect assertions I don't doubt that the publishers of these journals would be very unhappy about it, so please for your own sakes double-check these ">$9000 for a single article" assertions.
b) I checked the new, revised dataset upload
• Acta Palaeontologica Polonica – this journal is listed in DOAJ and has been listed (without interruption) in DOAJ since 2005: https://doaj.org/toc/1732-2421
Yet in their dataset the authors have this journal scored in their “Is DOAJ OA?” column as (blank), which I presume is “No”. Why?
The categorization of journals as wholly open access (or not) seems very confused, but perhaps this does not matter to the specific analyses presented in the manuscript?
I picked out
* Acta Botanica Cubana
* Bulletin of Zoological Nomenclature [open access at the biotaxa platform but behind a paywall at the BioOne platform] https://www.biotaxa.org/bzn
* Truebia
all of these I would call open access journals.
Lots of the cells were coded with "not 100% clear!" which I feel is a bit amateurish.
The "2019 subscription price (US$)" column remains conspicuously empty / unfilled with data entered for only ~86 journals.
Subscription journals included in the dataset like Bioinformatics (published by OUP) do have subscription list prices available online e.g. at https://academic.oup.com/pages/purchasing/institutional-journals-products/subscribe-by-title#price-list for OUP and here for Wiley subscription journals: https://onlinelibrary.wiley.com/library-info/products/price-lists (e.g. for Zoologica Scripta, a Wiley journal where no subscription price is given in the dataset)
It is a little disappointing that the data gathering for this column appears haphazard. I appreciate that journals are bundled, and sometimes non-disclosure agreements have been signed, but even so, why not just use the public list prices?
line 531-532: "three metrics of benefit to researchers (i.e., JIF, citation rate, journal usage by readers)"
I think this needs re-wording. Do you really mean to imply that Journal Impact Factor is 'of benefit' to researchers? Perhaps you meant 'of interest ' ? I'm not sure that the regime of Journal Impact Factor mania necessarily benefits research or researchers (on the whole).
Of the 3 reviews, while all are positive one delved deeper into the underlying data and found a number of errors which are a concern. This must be addressed in a revision, and as it should not be expected that the referee checked every example, then all the journal's OA policy and APC should be rechecked. All the referees make some additional constructive suggestions as well.
[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]
I have suggested some additional references that could be added in the introduction and Discussion sections; also a slight edit of Figure 3a.
Have made some suggestions - see document attached - to reword parts of the text for greater readability.
I have suggested some additional information to be provided regarding the statistical analyses performed.
Have also asked some minor points to be clarified in the methods - see attache document.
Have made some suggestions of additional discussion points - see attached document.
This is a well-written and well-structured study that will be of interest to the whole of the scientific community, especially environmental scientists.
I have made some minor comments / suggestions throughout the text, to provide more clarity / detail and improve readability.
Great effort by the authors.
In this manuscript the authors gather and analyze publication data on 34 out of 39 (over 87%) of the faculty of the Department of Ecology & Evolutionary Biology, at the University of Kansas (KU), as of March 1st 2020, to examine in which venues they publish, edit, and review-for, and the costs and prices associated with each of those publication venues.
I think this is really interesting work. I’m glad to see that knowledgeable KU librarians were brought into the collaboration to help with the work - similar research done by academics-only such as https://peerj.com/articles/7850/ has faced criticism and doubts over validity for not including librarians.
Key to this manuscript are the data, the software, and the exact processing/coding steps. I thank the authors for publicly archiving this in a suitable repository so they can be available for the reviewers and the public.
With regards to the underlying data https://kuscholarworks.ku.edu/handle/1808/32708 did the authors upload the correct file?
The data requires major revisions to make it robust, accurate, and clear.
In terms of coding “OA friendly” - I think a lot of journals have been miscoded, or else the coding methodology is insufficiently described to encapsulate what was done.
Take for instance the journal Nature , the authors score this as ‘0’ in terms of “OA friendly”. From the manuscript I presume ‘0’ translates to “unfriendly (i.e. author not allowed to post any post-review version of their paper)”. Assigning ‘0’ to Nature seems factually untrue to me. In actual fact, the journal Nature permits authors to make public their post-review accepted manuscript after only a 6-month embargo. See https://www.nature.com/nature-portfolio/editorial-policies/self-archiving-and-license-to-publish “When an article is accepted for publication in a Nature Portfolio journal via the subscription route, authors are permitted to self-archive the accepted manuscript on their own personal website and/or in their funder or institutional repositories, for public release six months after publication”
Likewise, the journal Evolution, is given a '0' in the author's dataset, but it permits authors to upload their post-review accepted manuscript after a 12-month embargo (according to Sherpa/Romeo). I think the authors need to have a look again at how they’ve scored each of these journals for “OA-friendly” and perhaps even consider a more sophisticated coding scheme to encompass the variation in length of embargo and other OA-related attributes? The coding scheme of 0,1,2 is arbitrary and unusual, it needs more justification at the very least.
Why are there unfilled/blank cells and what do unfilled cells mean for each column?
For instance, I don’t list all <blank> cells, just emblematic examples which need to be addressed:
a) the journal Molecular Ecology has a KU 2019 Subscription Price of <blank>. Molecular Ecology is NOT a fully OA journal, so the value in this cell ought to be >0 and not <blank>. If KU does not currently subscribe to Molecular Ecology then state that in the cell. Even then, I would have thought a representative price quote could still be sought and entered? Needs explanation at the very least!
b) Average (COUNTER) Usage of 2016 of PLOS ONE is given as <blank>. I think this ought to be coded more explicitly as e.g. UNKNOWN or DATA UNAVAILABLE rather than <blank>
c) In their dataset, the journal Evolution has an Article Processing Charge value of <blank> . Evolution is a ‘hybrid’ journal that permits authors to have their article open access if they pay an APC, the 'standard' APC at 2022-05-17 is listed at a price of $3000 USD. <blank> insufficiently describes the situation of this journal in respect to providing open access &/or APC pricing. Please always be sure to refer to APCs as prices and not costs throughout the manuscript. The publisher often sets the APC as a price they determine, and this price may be completely divorced from the actual cost of running the publication venue.
d) Why is there no COUNTER usage data for Phil Trans of the Royal Society B? It is another unexplained <blank> where I would expect there to be >0 data given this is a subscription/hybrid journal and the publisher does seem at first glance to provide COUNTER usage data.
e) Why is “OA friendly” <blank> for many journals e.g. Phil Trans of the Royal Society B? This is just pure error - there must be a categorization for each and every journal given the current coding scheme.
f) Please provide an ISSN or ISSN-L for each journal, the journal name alone is not a unique identifier. “Journal of Insect Science” may refer to either one of the following two journals:
Journal of Insect Science (Entomological Society of America)
OR
Journal of Insect Science (Indian Society for the Advancement of Insect Science)
g) Not all open access journals are listed in DOAJ. Actually a great number of open access journals exist beyond DOAJ’s listings e.g. read the The OA Diamond Journals Study - Science Europe https://scienceeurope.org/our-resources/oa-diamond-journals-study/
This is more than just a theoretical problem. It is a practical one e.g. Bruce Liebermann has an article published in The Open Paleontology Journal (Bentham, now discontinued). This journal is/was fully open access. At one point in time The Open Paleontology Journal was DOAJ-listed, but the journal is no longer DOAJ-listed. Which brings me to another point: what version and/or timepoint of the DOAJ-listings did you use in this analysis? The DOAJ-listings are not static. It would be good to include some versioning information for reproducibility. If any journals were manually determined to be fully open access journals (outside of DOAJ-listings) perhaps state that? Pakistan Journal of Biological Sciences is one such title inc in the scope of this EEB analysis that to my eyes (looking at the journal website) seems openly accessible to readers, with articles distributed under CC-BY, yet it is not DOAJ-listed. Oddly "Pakistan Journal of Biological Sciences" is also only listed as "1" for "OA-Friendly" which must be an error given the final pdfs are open online under CC-BY?
I can't really trust this article until the underlying dataset is properly re-scored and more fully explained.
It would be nice if the authors wrote some nicer words about the value of research & scholarly writing that is NOT published in peer-reviewed journals. The way in which the 'not peer reviewed' or 'not published in a journal' were abruptly discarded from the analysis is troubling for the overall aim of the manuscript to look at what and where scholars and institutions are investing their time in supporting. Journals aren't the be all and end-all of scholarly communication. Monographs, book chapters, even blog posts can be excellent contributions to research, and all can be/are cited too.
What about selection bias? I note only 34/39 faculty were included and that the authors "requested they create [GS] profiles"
I think you should discuss the weakness of this design a bit further. How many of the faculty specifically created a GS profile at the authors request? Why would an EEB faculty member actively not want a GS profile and may people like this (anti-Google?) have a different publication, editor, and review profile?
What about the completeness of each of the 34 GS profiles included? No mention is made that GS profiles are malleable and owners of each profile can include/exclude publications on their GS profile at whim. I think you may want to consider evaluating this weakness, perhaps with reference to another data source?
Likewise, what is the completeness of annual reports submitted to the department detailing editorial handlings and peer review reports? If a faculty member was in a rush to submit their annual report, might they not accidentally forget to include an editorship or a peer review done here or there? Perhaps you can cross-check with Publons for one or more faculty members to help assure the reader of the completeness of this data?
Please include a link to an archived page as close as possible (in time) to the time-point at which you measured the number and identity of EEB faculty at KU (1st March 2020). I found a snapshot of the page at the Internet Archive from December 2019 here: https://web.archive.org/web/20191206051920/http://eeb.ku.edu/faculty but if you have any other/better/closer archived snapshots to prove that assertion then please use them.
Line 158 “high-ranking” high-ranking according to who or what? I disagree with the concept of ranking journals (see various works by e.g. Bjorn Brembs). When I do rank journals e.g. in a spreadsheet I tend to do it alphabetically, so American Naturalist is 'high-ranking' for me :)
The analysis fails to acknowledge explicitly that lots of peer-reviewed journals have specific sections within them that are NOT peer-reviewed. How was this accounted for, if at all? Thus an author may have published an item in Nature, but that item might not necessarily be peer-reviewed. Were 'editorials' , 'reviews', 'letters of concern' , 'erratum notices' , 'corrigenda' and other such exotic article types all counted the same as typical 'primary research articles' ? It might be good to remark on the effect of this if so in a 'limitations' section of this manuscript.
How representative is the EEB department at KU of any other EEB department at any other university? Some words to address this question might be nice?
Please consider more clearly indicating that Clarivate’s Journal Impact Factor (JIF)™ is a proprietary number. Specifically, capitalise all references to it AND include the ™ symbol to remind readers that it is also trademarked in addition to being untransparent, negotiable, statistically illiterate [1], irreproducible, and proprietary. Do not refer to merely ‘impact factor(s)’ – as such lower-casing might imply a generic, reproducible, statistically logical calculation. Also be more specific than just describing it as an 'average' (e.g. line 258). Clarivate famously calculates the mean rather than the median, which given the known highly skewed distribution of citations across articles in a journal is a statistically illiterate practice [1], hence you should tell readers it is the mean (which is statistically inappropriate) not the 'average'.
[1] Stephen Curry (2012) http://occamstypewriter.org/scurry/2012/08/13/sick-of-impact-factors/
The findings are not valid until they fix the dataset. The given dataset has a LOT of errors and unexplained cells.
Despite everything I've said above, I do think this manuscript presents some very interesting research and I would be keen to see it get published in a peer-reviewed journal. I merely wish to help make it more robust before it gets published in a peer-reviewed journal.
The calculus of investing time, labour, and hard cash in publication venues is an extremely important thing to examine and I congratulate the authors on gathering this data to start doing this, in the open so that others can see, think and learn from the results.
Please see additional comments.
Please see additional comments.
Please see additional comments.
I was pleased to review “Relationships among cost, citation, and access in journal publishing by an ecology and evolutionary biology department at a U.S. university”. The topic of the paper is important and suitable for PeerJ. I have a few comments for improvement, which the authors should have no trouble addressing.
I was pleased to see that the authors archived the data and custom script used in this study. However, I would encourage the authors to also share the script used for the statistical analyses (test of relationships between variables) and to create the figures, as well as to improve the dataset by providing metadata (e.g., a comprehensive description of the column headings, units and abbreviations) in a separate readme file and applying tidy data principles (no blank cells, one value per cell [i.e., no units in cells]) to facilitate reuse (White et al. 2013).
Relationships between variables were examined by means of regression analysis, which assumes that one variable is dependent on the other. This dependency is not straightforward in almost all the relationships examined and, as such, the use of a correlative approach is more appropriate. The authors should consider reporting correlation coefficients rather than coefficients of determination. P-values should also be reported to three decimal places rather than above or below a critical value (Head et al. 2015).
The three paragraphs from L257-285 seem out of place and/or weekly linked to the text in this section of the manuscript. The content of these paragraphs is fine (although it could be condensed) but links to the discussion of the results are unclear. Consider rewriting, moving the text to another section, or drawing tighter links to the results discussed in the text of this section.
Numerous references cited in the text do not appear in the bibliography.
L55-64. Consider citing (Mekonnen et al. 2022) and (Larios et al. 2020).
L77. Consider defining ‘journal use’ on first use.
L79. Comparator needed in this sentence.
L172-174. This sentence seems more appropriate for the discussion.
L185. Consider specifying EEB publications at the University of Kansas.
L232. Consider referring the reader to the relevant figure.
Figure 3. The caption mentions publication frequency but the y axis in panel A is number of EEB papers. Please specify whether this is a frequency or a number and, in the latter case, the time period covered.
Regards,
Dom Roche
References
Head ML, Holman L, Lanfear R, Kahn AT, Jennions MD. 2015. The extent and consequences of p-hacking in science. PLoS Biology 13:e1002106.
Larios D, Brooks TM, Macfarlane NB, Roy S. 2020. Access to scientific literature by the conservation community. PeerJ 8:e9404.
Mekonnen A, et al. 2022. Can I afford to publish? A dilemma for African scholars. Ecology Letters 25:711-715.
White EP, Baldridge E, Brym ZT, Locey KJ, McGlinn DJ, Supp SR. 2013. Nine simple ways to make it easier to (re) use your data. Ideas in Ecology and Evolution 6:1–10.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Excellent Review