bionieuws 5 | 14 March 2003 | volume 13 | pp. 8-9

Still early days for Web publications

Escaping from the grip of publishers, reviewers and hype

Have paper journals had their day and does the future belong to the Internet? Or is the power of leading journals unbreakable? On PDF files, impact factors and comparing apples and oranges.

Thanks to a donation of $9 million, the non-profit organisation Public Library of Science, PLoS, is going to launch two journals for and by scientists, PLoS Biology and PLoS Medicine, later this year. The intention is that these journals will be freely available online to everyone. Authors will contribute to the financing of the journals by paying for each article they publish, but they will retain the copyrights to their papers.

PLoS's aim is to free the publication of scientific papers from the yoke of commercial publishers. These publishers make a lot of money from expensive journal subscriptions, while scientists supply the contents. But scientists cannot do without the publishers, as publishing is a scientific necessity.

One Internet journal that has already thrown off the yoke is Conservation Ecology. Papers in Conservation Ecology are not printed, they are available from the journal's website as PDF files. Just as will soon be the case at PLoS, the papers are again freely available to everyone, while the journal derives its revenue from the payments that authors must make if they wish to publish a paper.

Bruno Ens, a behavioural ecologist at Alterra/Texel, is a member of the journal's editorial board. He understands the advantages of a purely electronic journal. 'I myself have noticed that I am increasingly averse to paper,' he says. 'I find it very easy to save offprints on my hard disk as PDF files. I can look up articles more quickly than in paper files. And it also makes a difference to the trees, but whether that is offset by the extra electricity you use...?'

In other respects, publishing his colleagues' work in the journal has fallen somewhat short of the behavioural ecologist's expectations. The impact factor is not high enough, it costs money to publish an article and digital publishing is still not quite the thing. 'If it is not printed, it somehow seems less real,' explains Ens. 'But it's my bet that it's just a question of people getting used to it, although it is taking longer than I thought it would.'

Conservation Ecology has been in existence since 1997, but has found few imitators. Perhaps fully digital publishing is a special niche. In this case, the online journal is published by The Resilience Alliance, an organisation that wanted to be fully independent in order to spread its message that 'ecologically, things are going very badly for the planet', in Ens's words. 'That's something I wholeheartedly support. I also think it's a good thing that precisely this sort of science - at the interface between ecology, social sciences and public policy - is available to everyone.'

Cherry-picking

The online review system Faculty of Thousand, F1000, is not free, but it is digital. It does not publish anything itself, but cleverly takes advantage of the overwhelming quantity of literature. Impact factors already assist in the selection of journals, but there are still many left over to be read. Subscribers to F1000 are pointed in the direction of eye-catching publications by scientists. Each month more than a thousand biologically-oriented scientists assign a rating to a paper that has impressed them: recommended, must read or exceptional. The paper is then given a label: Novel finding, Technical advance, Interesting hypothesis, Important confirmation or Controversial findings. Naturally, the papers are also subdivided by content. For example in the Biochemistry Faculty, Protein folding section or the Microbiology Faculty, Innate immunity section.

F1000 is a peer review system that lets experts pick out the best "cherries". Impact factors do indicate - as it were - on which trees most of the best cherries might be found, but a lot of them are still hidden from view. F1000 tries to bring all the cherries to the fore. The virtual faculty does that quite explicitly with its category Hidden Jewels. This contains recommended papers from less widely read specialist journals which most scientists will therefore have missed.

One of the thousand scientists who help to track down interesting papers is Sybren de Hoog of the Fungal Biodiversity Centre in Utrecht. He is one of the more than twenty Dutch participants in F1000, and he is enthusiastic: 'It's pretty incredible, but it works well. It's a sort of Reader's Digest. You can select papers that you yourself think are good - from any conceivable journal. In fact, they encourage you to unearth unusual things.' Because his specialist field - fungi - is not so well known, several of the papers he has recommended have already been promoted to Hidden Jewels.

The intention is that the more than one thousand members of F1000 should recommend two papers a month. De Hoog: 'Because I have kept to that so well, I have already earned a subscription for all my colleagues. It's also nice that you can recommend labs from developing countries for a free subscription.'

Hype

De Hoog considers F1000 a welcome addition to the existing range of journals. The attention paid to high impact journals has serious disadvantages, he explains. 'You create hype, and everyone runs along behind it. The threat of bioterrorism, for example. But in the meantime a lot of money flows towards it, so then you actually do have to participate in it. Other examples of hype? Well, our neighbours, the Hubrecht Laboratory. Not really hype as such, but it does very well in the market.' De Hoog is referring to the RNA interference research by Plasterk and colleagues.

Ton Bisseling, a molecular plant biologist at the University of Wageningen, has been invited to take part in F1000, but he is not yet active. He is, however, active on behalf of Science, helping the editors to select from the submitted papers. He gives his opinion on around 100 to 150 articles a year, indicating whether or not he thinks the research is innovative enough for Science. The subjects vary from molecular to ecological, but are limited, at Bisseling's request, to the plant sciences. He gives his comments a figure between zero and five. 'That indicates to what extent I feel sure about my assessment.'

Bisseling does not believe that the attention paid to high impact journals such as Science has an adverse effect on specific specialist fields. 'If there is an important development in a specialist field, then those results will appear.' In ecology, for example, he considers that the recent feedback between knowledge of plants in model systems and field tests has received the attention that it deserves.

Logically, Bisseling also thinks it right that researchers should be compensated according to the impact factors of the journals in which they publish. 'It has a steering effect on finances.' In this way, subsidies go towards innovative research.

Stalinism

Someone who disagrees with the regard given to high impact factors is the Utrecht ecophysiologist Hendrik Poorter. He considers it to be 'Stalinism at the top'. 'You stand on the platform with the important people, so you are important too.' But that is not necessarily the case, explains Poorter. An analysis he carried out showed that half of all papers more than ten years old had been cited less than ten times. That applied both to leading journals and to more mediocre ones. Many papers thus simply gather dust, even if they sometimes have a golden border because of the elegant journal in which they are published.

Wim de Leijster, the secretary of the Experimental Plant Sciences Graduate School (EPS) endorses that view. 'It doesn't matter how brilliant a paper you've written, if no one cites it, then it's been a waste of time.'

De Leijster carried out a citation analysis of EPS scientists. Apart from the level of publication, De Leijster also attached weight to how often a paper was cited. 'There does not exist a 100% correlation between the level of publication and being frequently quoted. Even a paper in Plant Journal is not always cited.'

In the analysis De Leijster had to factor in differences in citation culture between specialist fields. The impact factor of a journal is calculated by dividing the number of citations to papers published in that journal in the two previous years by the total number of papers published in the journal in the same period. If the impact factor is high, there have been more frequent references to papers in that journal. The higher the impact factor, the greater the influence of the journal. The rate of circulation and the intensity of citation are, however, much higher in the biomedical or molecular fields than, for example, in ecology. 'Personally, I try to see whether researchers are reaching the best journals in their own area.'

Poorter adds a critical note in the margin of such analyses. 'Take for instance the publication value that my graduate school calculates. The impact factor of the journal is divided by the number of co-authors of a paper. If I publish an article in Nature together with nine other people, that gives me three points: almost thirty divided by ten. The highest ranking ecological journals have an impact factor of 3 to 4, on average it is around 2. That means that I am better off tagging along with an article in Nature, than spending a year doing my own ecophysiological research and publishing it in a high-ranking ecological journal.'

The emphasis on impact factors thus has an adverse effect on the ecology, asserts Poorter. He illustrates this with the example of a colleague who works at the Flemish Institute of Biotechnology in Ghent. He is expected to publish only in journals with an impact factor of five or higher. Poorter: 'For his colleagues who do AIDS or cancer research, that is perhaps a realistic requirement, but for him as a plant biologist it leaves only three journals he can publish in. Managers think it an excellent idea to assess their employees according to high scores, but this is really a very bad development.'

Poorter concedes that within a specialist field the impact factor of a journal is a good indicator of the quality of papers. 'But comparing AIDS research with work on tulip bulbs is like comparing apples and oranges. You can't do it.'

Rinze Benedictus

bionieuws 5 | 14 March 2003 | volume 13