Arguments
Deciding who should pay to publish peer-reviewed scientific research
Posted on 18 September 2014 by John Abraham
There is an important discussion to be had about the future of scientific publications.
As a practicing and publishing scientist, I am judged by the quality and quantity of my contributions to the scientific community. Traditionally, this comes down to counting how many papers I publish and weighting them by the quality (impact) of the journals where the papers appear. A fancy word for this is “Impact Factor,” which is a measure of the frequency papers in a particular journal are cited compared to the number of pages a manuscript is.
The highest impact journals are often the hardest to get published in, sometimes having acceptance rates as low as 10%. Typical impact factors depend a lot on your field of study. In journals like Nature and Science, the impact factors are very high. In specialized journals and in specialized fields, the impact factors are much smaller.
In my native field of heat transfer, impact factors as high as 2.5 are rare. In climate science, flagship journals like the Journal of Geophysical Research and Geophysical Research Letters have impact factors in the 3–5 range – this means that the technical field of geophysics has a higher citation rate then say, heat and mass transfer. Journals such as Nature and Science, broad-category journals with huge readership, have citations rates of 42 and 31, respectively.
In this traditional model, universities pay each year (often thousands of dollars) to carry the journals. The universities then typically received both hard copy and e-copies of papers which faculty can then obtain. More recently, many library consortia have gone to an electronic-only system. It is probably obvious that with strengths of this system come weaknesses.
A glaring problem is that the subscription fees are quite large and very few practitioners in a field purchase the journals. Instead, they can purchase specific papers that they are interested in, often for $20–40 per article. It is commonly said that the papers are “behind a pay wall.” But, this pay wall is important.
A publisher cannot simply give papers away for free – they would rapidly go out of business. On the other hand, an author can opt to make their papers available without a pay wall, but the author has to pay for this option. My colleagues and I recently wrote a major ocean heating paper and paid multiple thousands of dollars to make it freely available. This money came from our research budgets – budgets that are already tight.
So into this mix enter open-access publishers. Instead of selling papers, they make the articles freely available to the public. On the one hand, this system dramatically alters who can gain access to articles. The papers can be freely downloaded anywhere in the world (hugely important if you are a researcher in the developing world). In addition, open-access journals typically do not print papers in hard copy form, thus saving money on printing and shipping. But how can these journals survive? They do that by charging the author. Fees range anywhere from $100–$1000 or so.
So, whenever a scientist opts to make their papers open-access, they (or their institution) are paying for this service. It is important for the public to recognize this. Publishing behind a “pay wall” does not mean a scientist is hiding anything – it is a necessary part of the business model of traditional journals. And, when journals have a pay wall, they are not gouging the public. These payments cover the costs of editing, printing, formatting, etc.
Many journals, whether open access or not, are for profit enterprises, but they are enterprises that provide a valuable service to society. (Other journals are published by universities and consequently do not fit into this mold).
A problem can arise, however, when open-access journals solicit papers and publish them with sub-quality review as a means to increase profit. These types of journals are often termed “predatory” in that they prey on professors (although faculty shouldn’t be so naïve).
All of this presents researchers with a tough choice. Publish in a top-tier journal and likely get a higher profile amongst your peers but a significantly reduced readership outside of academia. Or, publish in open-access venues and increase your public readership but incur publishing costs and lower citation rates among your peers.
I would say that most academics have chosen the first approach – shunning the new open-access route and staying with the tried and true journals. I am more mixed. I think there are real benefits to making my research available to a larger audience. On the other hand, I want to ensure that my work is well received by colleagues and has an archived life so that years from now, researchers can read my contributions.
In the end, most of my work is published in the very top thermal-sciences journals such as the International Journal of Heat and Mass Transfer, and Numerical Heat Transfer, with impact factors 2.5 and 1.8, respectively. Recently, I have written in open-access journals, particularly when the subject is not specialized, or when I think it will be of interest to a broad audience. When I write in open-access venues, I hope that my publication payments will guarantee a permanent archive, a broad readership, and top-quality editing.
With many new journal ventures beginning each year, it is hard to anticipate which will be around in the years and decades to come. Some ventures are clearly going to be successful, such as Hindawi, Copernicus, SciELO, and Scientific Research. Others may not be successful, particularly if they cannot ensure high quality articles within their pages.
An expert on this topic, Dr. Lars Bjørnshauge has taken an active role to categorize these open journals. He founded the Directory of Open Access Journals in 2003 which applies quality metrics to the venues. He summarized his view when he told me,
Open access publishing (with or without) Article Processing Charges is here to stay, not only because of the well documented benefits for the researcher, but as well because more and more research funders, universities and governments has realized that open access to public funded research benefits research, industry, innovation, health etc. and therefore has issued open access mandates and policies requesting that the results of the research should be freely available and reusable. People can learn more about open-access mandates at roarmap.eprints.org.
The DOAJ provides a global service listing and disseminating information about fully open access journals providing extensive information about charges, licensing, archiving arrangements etc. The new criteria will enable researchers, universities, research funders and other stakeholders to find good publication channels.
So perhaps it is best to share my own guidelines – questions I ask when I decide to go the open access route.
I'd like to see a system were non profit journals get grants and then go all open access. Now tax payers and college tiotion money goes to libraries and then to journals in sub fees. Just fund journals directly and have open access for all.
John Abraham, recently during discussion of a very poorly referreed article by Ross McKittrick, I had reason to look up Jeffrey Beal's list of predatory pubishers. The journal that published McKittrick's paper was part of the Scientific Research group, and indeed, Scientific Research appears as number 405 on Jeffrey Beal's list. You, however, recommend it, or at least indicate it is going to be a successful academic publisher. You may want to reconsider that judgement, or alternatively give your reasons for disagreeing with Beal.
Mike, such journals already exist. One of the top journals in my field is the Journal of Machine Learning Research. This runs on volunteer effort (most reviewers and editors for commercial journals are also unpaid, so that isn't much different). The difference is the journal is published online and the papers are open access and neither the authors nor the readers pay. A print-on-demand version is made available for libraries that want it. It is not completely clear that there is any need for commercial journals, given the success demonstrated by JMLR.
HOWEVER, I suspect the success of JMLR is largely due to the support of the very top researchers in the field that form the editorial board and panel of action editors etc.
Most journals, even the commercial ones, tend to allow authors to distribute pre-prints of their papers from their websites or institutional websites, so the papers are still available to the public. You can find out the journals policy using the Sherpa/Romeo database. However, if all else fails, sending an email to the corresponding author asking for a pre-print of the paper will usually be successful, most authors are only too pleased to hear somebody wants to read their work!
If taxpayers are paying for a researcher's research, then taxpayers should have open access to that research.
The concept of for profit journals is as egregious to me as for profit hospitals. Noth have an obvious conflict of interest.
Kernos... You do have access to the research, through a subscription. There happen to be costs that come into play regarding the publishing and distribution of research. That's what subscriptions pay for.
Perhaps you would be in favor of raising taxes to underwrite those costs so all research can be made freely available to the public.
A more efficient approach would be for the research funding councils to fund their own high quality journals, so the publication charge need only cover the bare costs of publication. The journal I mentioned earlier (JMLR) apparently only costs a few tens of thousands of dollars a year to run, which shows this can be done. Then tax payers money would not go to commercial publishers profits, and could be spent on more science.
I have no objection to people wanting to make money, but at the same time, I do want to see taxpayers money spent wisely.
There is a method to determine the ranking of chess players that might work here. http://www.chess.com/article/view/chess-ratings---how-they-work
We can assume that all places of publication are the same since everything on the internet is, for all practical purposes, in the same place.
It will require some refinement since there is not likely to emerge as objective a method as who has won the most games against other players. One could possibly get google could add their expertise in ranking things. The strength of a ranker might start off with some combination of Years of post graduate education, number of papers published, number of patents, prizes won in the relevant area. There might be a table constructed of relevancies. Other factors might be the global ranking of the unuversity or institution for which the author / critic works or derives primary income.
Surely with all this intellectual horsepower around the people who care could come up with come consensus as to how this would work, at least to start. And then maybe once every couple of years, revise the system. Again, you may wish to start by asking google to advise.
think of it as a mashup between wiki and google GW Call it sciub. A person posts a paper to scipub. It gets reviewed and that review goes at the bottom of the paper along with the ranking of the reviewer who also ranks the paper. The first 5 reviews are posted in order of their publication, after that they are listed in decending order by the ranking of the reviewer, but can be re-sorted by date, authors name, institution with whom author is affiliated -— other stuff.
let me add that any person who thought a paper had value could donate some money to the author, his listed projects, his institution etc.
Maybe the means of calculating the impact factor itself could be changed to encourage open-access publishing. This might encourage researchers to go to open-access journals first. Perhaps an open-access journal could justify having more IF points because we might assume that its articles are more widely read, since there is no barrier to access from people with no instititional library access. It's not just citations that matter, but eyeballs, also.
We have already paid for the research. We shouldnt have to pay again. With most journals online, costs are minimal anyway. If it has to come out of the grant money so be it. Include it in your grant proposal.
If the public is to become more science literate, access must be available to all. $20-$30 a paper is impossible for most Americans. And, avoid publishing in for-profit journals. They need to disappear.
Kernos... In fact, not all research is paid through tax funded government grants. A large amount of research is paid for through charitable trusts. So, you don't really have any idea which research is publicly funded. You're merely assuming you "should" have free access because your tax dollars paid for it.
I can guarantee you that free access to research is not going to do anything to improve the general public's scientific literacy, since nearly all research is written on a level that requires a great deal of knowledge in any given field.
Educating the public on scientific issues is a great thing! But it's an issue entirely separated from access to research.
Kernos @10, you have not already paid for the research. In addition to the fact that not all research is paid for from taxes, as noted by Rob Honeycutt, not all research is done in the United States. By your argument, only those articles whose research was fully funded by the US government should be available to you (and other US citizens), and such articles should be available for free only to US citizens.
Further, even as regards to the US funded research, what is funded is the research - not the publication. When you can persuade your congress to set aside additional funds so that all tax funded research attracts an additional grant for public access publication, then your claim would be valid. Until then your interest as a taxpayer gives you no more right to free access to the publications than it does to free access to CIA reports.
I will grant that you may have a right to FOIA the paper, but doing so will just result in the agency through whom you initiate your FOI request purchasing a copy of the article, adding their own administration fee and then billing you for the total as the cost of satisfying the request.
When research is fully or partly publicly funded, any publications arising from it absolutely should be freely available to the public. Publication costs should be part of the funding, since somebody has to pay for it and it may as well be the grant payers rather than university libraries.
Of course, there is publicly-funded research information that cannot be published, for example, on weapons or other issues of national security, but cost-free public access to this is beside the point, since such research is not published anyway. If the CIA were to write articles to be published in a science journal, then those papers ought to be free to everyone to read, but nobody, surely, is demanding open access to all of the CIA's reports.
Of course, there is a good case to be made for corporations to be allowed to keep their own research proprietary, even if it is partly funded through tax incentives. Again, we are not talking about free access to information that is not published.
The lucrative business models of the for-profit publishing houses get a free ride on the backs of taxpayers and, in the process, restrict public access to knowledge. This needs to change.
Andy Skuce @13, I agree that free public access to scientific research would be very desirable. I disagree that that desiderata is a right based on tax payer funding. That becomes obvious not just from comparisons with CIA documents, but with government information generally. Even data which is publicly accessible by legislation will still attract a fee based on costs of accessing the data in FOI requests. Some people may consider that a bad thing, but the alternative is to allow government departments (or researchers) to be paralyzed by frivolous FOI requests that are paid for by the department or researcher. A $35 fee for a paper is a very reasonable value relative to that for most FOI requests.
Nor, does government funding transfer intellectual property rights. That is, the intellectual property rights on the research remains that of the researcher - and ought to do so. You tacitly acknowledge this in making an exception for publicly funded research by private corporations.
If we think free public access to research is a good thing, then we need to get our representatives to set up and fund institutions that provide that service. Personally, I think that should be done by funding public accessable journals, with grants to such journals based on their impact factor (or impact factor relative to specialization) so as to discourage the development of a vanity press in science. Further, the grants should be sufficient to provide a small stipend to editors and reviewers based on the impact factor of published papers (ie, rewarding quality over quantity) to attract the highest quality editors and reviewers to the public access journals. Further, existing private publishers should be invited to transfer there journals to a public access model on the same funding basis as journals set up specifically for that purpose.
What ever the model pursued, however, we need to recognize that there are issues relating to international cooperation, and potential pitfalls. We also need to recognize that this is a new service from government that we are requesting, not one that we already have a right to.
Tom Curtis Your rather hard line stance on journal publication is likely to enhance rather than diminish the proliferation of open access journals. I'm not sure if you have ever applied for a grant but if you put in a line requesting funding of, say, $1000 for future publications. you might well jeopardise or at least feel you are jeopardising, your chances of obtaining a grant as money is very tight and grants are very competitive. As an academic who applies for reserarch funding, I can assure you that an application has to include a substantial portion of funds requested for the University and very rarely does this portion cover publication costs. But Tom Curtis, Universities wax fat on the efforts of researchers as their success in attracting grants adds to the public profile of the university. This in turn. attracts more students and thus leads to more income. Similarly journals wax fat on the research papers they receive and for which they levy a charge on the researcher as without these papers, they are as nothing. So Unis charge researchers so they can attract and charge more students and journals charge researchers so they can attract and charge more readers Can you not see what an invidious system it is and why open access journals are regarded by the "establishment" as the "spawn of the devil"? Sure, looking at the publishers of open access journals lambasted by Jeffrey Beall he certainly has a point. However to condemn open access because at the moment it is far from ideal is not the best approach at all. It would be better to get these journals to meet the necessary standards. A way this might be achieved is to adopt/adapt the approach of the mass media to support the costs of on line publication such as judicious use of advertising. As an aside, is it pertinent or indeed impertinent, to ask if bringing SkS to readers incurs charges and if so to ask how these charges are met?
Ashton, as to your final question, the answer is here.
Most journals these days allow authors to self archive (i.e. put on their website) a pre- or post-print of the paper. So even with commercial publishers, it is not generally the case that publically funded research findings are not freely available to the public.
The next point is that the public gets for its tax dollars only what it pays for. If society decides that they want papers/data to be freely available, then that is an extra cost and they need to pay some extra tax to cover it. This is rather like the problem with the availabilty of national weather service data. The data has commercial value, so governments generally decided that national met offices should exploit this commercial value to cover some of the costs of its collection and so reduce the burden on the taxpayer. Now people are demanding free access to the data (not that they actually do anything useful with it), which means that the tax payer must now make up the shortfall in the met offices income caused by not being able to commercialise their data.
RCUK make the funds available via a block grant to each university, so it doesn't actually appear as a cost on grant proposals anyway (I am not convinced that is a good idea).
As regards predatory open access journals, the UK's Research Excellence Framework (REF) grades research output into five gradings from "four star" (Quality that is world-leading in terms of originality, significance and rigour) through to "one star" (Quality that is recognised nationally in terms of originality, significance and rigour), and "Unclassified Quality" (falls below the standard of nationally recognised work). Work published in such journals is unlikely to get a three or four star rating, and hence there is a big disincentive to publish there, as the funding is likely to be non-linear with proportionally more funding directed at the departments with a higher average grade.
If you want to see how open access should operate, take a look at JMLR.
Ashton @15, I don't know which Tom Curtis you are responding to, but it is certainly not me. I suggest you read my post @14, which you probably did not see before posting your comment. It makes it quite clear, however, that you are arguing against a strawman. Specifically, I do not disagree with open access. Indeed, I am strongly in favour of it. What I disagree with are specious arguments that we, as taxpayers, are already entitled to open access without having made specific legislative and funding provision to support it.