Thursday, June 20, 2013

Open Access - solid reasoning, or a bunch of hypothetical statements?

This post will be the first of a pair triplet on open access journals, which offer author's the ability to pay the publication fee in exchange for their article to be free to anyone wishing to read it. Here I will discuss whether open access stands up to its claims, and my mum will respond in the next post using highlights from Scholarly Kitchen, a publishing blog she contributes to. My grandfather, who is well-versed in science communication, will finish up the trio with his comments.

So, what is open access and why are people so obsessed with it? 

Scientists are increasingly jumping on the open access bandwagon for two major reasons, I think. The first (lesser-cited one) is that if journals are free and so more people can access them, that article is more likely to be cited, and more citations means more fame for the author and the institution, and possibly more funding (or accelerated tenure). The second reason is that the public, as well as scientists at smaller institutions, in under-developed countries, or independent scientists, should be able to have access to research without paying exorbitant fees.

Perhaps the greatest advocates for open access are the greatest consumers of knowledge - university libraries. These are the libraries which provide access to the literature upon which their most prominent researchers build and bring fame and money to the institution. However, these libraries face a dilemma - they must provide researchers with these articles, but the subscription costs for many journals are increasing faster than library budgets. Therefore, as long as open access journals offer the content researchers need, libraries have an opportunity to reduce costs. But these costs are then just diverted elsewhere in university budgets to pay for publishing articles. A quick Google Scholar search looking for "University of Massachusetts", "journal" and 2012 returns 16,000 articles published last year. If all of these were published in PLoS ONE, the largest open access journal in existence, then this would cost the university system $8,000,000, which is about one and a half times UMass Amherst's library materials budget. A saving to the university? Not quite. See, the fundamental issue is that library budgets have been a shrinking part of university budgets for the past 14 years, while money is diverted (at least at UMass) to buildings, bouncy castle parties, and other shows of grandeur which do little for learning and/or safety.
 
My new lab building to be at UMass. It instills an incredible sense of guilt associated with unnecessary destruction of historic buildings and materialism-driven construction of the new every time I look at it. Picture from mass.gov.



Returning to the idea of free articles being more available to the public, we must then ask why the public would want to be on the cutting edge of research. Many of these articles are difficult to understand, even for scientists, and the US government is moving towards making articles stemming from research it funds must be publicly available after one year (though I am unsure whether this is from the online pre-publishing date, which may be a year before the official online or paper printing date). There are clearly many potential customers for these free articles - imagine a mother looking for alternative treatments for their child's illness, or an environmentalist group looking to advise the government on climate policy - but it is difficult to determine how much making an article open access increases the audience size and breadth. It would be interesting to compare view counts for "normal" and open access articles in journals with publish a mix of both and see if the open access articles receive more views. Of course, it is possible that people who believe their article is more pertinent to the public may choose to write and fund their article in a way which will optimize public viewing, thereby biasing view count results. But a quick idea might be nice - anyone want to do that study?  

As a final point on the Open Access discussion, I would like to briefly mention that in theory some researchers are worried about publishing in Open Access journals because the quality of editing is lower. I don't necessarily think that the quality of editing is lower, but it is somewhat perverse that basically researchers are paying to have their research published, and only those with money can afford to do this. This isn't much of a worry for lab scientists, which often work with very large budgets, but physicists and many other disciplines work on much smaller budgets that cannot easily absorb a $500 publishing fee. That said, there are arguably other editing rules which potentially do greater harm to the literature and knowledge base than who has the most money - for example results in which no change or difference was observed between two groups are basically unpublishable. And PLos ONE takes the view that ultimately each reader should decide on the validity of an article for him or herself, and demonstrate support for work by citing it. However, if a journal like PLoS ONE produces articles with public-friendly descriptions of the research, and the science has fundamental technical or ethical flaws, they may not be able to detect the errors, and, becoming more like News of the World than the New York Times, bad science can spread through the unfiltered public conscience and become a social fact. Of course, poor editing is not limited to open access journals, but the fact that they are so openly available means that editing needs to be stricter for freely-available than subscription-only journals, I think.

What do you think?

6 comments:

  1. Great post, Grace! I'm wondering what you mean by "editing," though. Do you mean peer-reviewing? Professional editors making decisions? Copyediting? PLoS ONE goes through the first two, but not the latter.

    I wonder though -- here I'm being especially (and intentionally) provocative -- in a DIY publishing world why we need journals at all. What say you to that?

    ReplyDelete
    Replies
    1. Thanks! I mostly meant peer review, but also a bit of professional editors and copyediting too. That is based on my one experience with a non-PLoS open access journal, though. That is not to say that the most reputable journals haven't published a fair number of articles which have or should have been retracted - or at least severely edited - to account for falsified data and/or sweeping, unbacked conclusions. Or that their copyediting is even vaguely effective (It no longer comes as a surprise to me if there is an unlabeled axis on a graph in GCB, or if the units are impossible). I guess the answer is that scientists need to take greater responsibility for their writing - I was very surprised after my PI submitted a paper with obvious typos and said that it was OK because the copyeditors would pick up on them. Even though everyone we interacted with at the journal had very poor English.

      As for why we need journals, I think there are a few answers. First is that we are trained to do keyword-driven searches to home in on exactly what we want to know, but this makes us miss out on large chunks of knowledge we might not think was related until we stumble across it in a table of contents. There are of course various "articles like this" tools, but they aren't smart enough at predicting what I might be interested in. The second reason is that, as much as the science should matter more than where it is published, there is a dominance hierarchy of journals, and the more top tier journals you publish in, the faster the track to tenure. The third reason is that I think people like the fact that if they want to know what is cutting edge in the world, they can go to Science or Nature and read up on it. So I guess that would be journals are needed because we trust the judgements of others to determine what is important - it is like paying someone to open and sort through the mail.

      Delete
  2. On the subject of PLoS One and editing/peer review, they have peer review for technical and ethical accuracy, like reputable closed-access journals; however, they don't make decisions based on novelty or impact to the field, unlike most other (society or what have you) journals. So interdisciplinary or less "sexy" papers that are still based on good science have a better chance of getting published. On the subject of fees, it is often possibly to get some or all of an OA fee paid by the department or another institutional office; and, for less-well-funded research and for research, from, say, developing countries, PLoS and many other OA journals will often reduce or waive the fee if you explain your situation to them.

    ReplyDelete
    Replies
    1. Ariellicle! My point was more that the cost doesn't disappear - it just moves elsewhere as budgets evolve. The amount of overhead on grants is already completely and utterly ridiculous in some departments, and I think more publishing in open access journals will just increase it. Unless funding agencies start imposing strict limits on overhead, which I think they should.

      Delete
  3. One issue that I'd be concerned about if funders start imposing limits on overhead, is that this would push researchers into publishing in the lowest cost journal which may or may not be the most appropriate. I don't personally think funders should be dictating where a researcher publishes her/his findings.

    ReplyDelete
    Replies
    1. I don't know they would always want the cheapest...I think they too want the best bang for their buck balancing journal impact factor (whatever that metric is) along with cost.

      Delete