Lately I’ve been doing lots of literature searches. Pulling out old papers I’d skimmed through at the start of my PhD and trawling them for information. In doing this I noticed how much peer-review journal publications have changed in the space of about 10 years. A lot of the articles I’ve been using are from the 1980s and ’90s, when paper copies of journals were still prominent. The paper itself was a solo star; submitted, reviewed and accepted in an unseen process before publication. All you had were the data held within the publication itself.
With the rise of the internet peer-review went digital and without the limits of a paper publication journal could include far more information with each article published. Many journals now allow supplementary information, tables and data to be uploaded along with the paper.
In the majority of cases I imagine this can only improve the quality of publications and data access. My personal literature trawl aimed to pull out other scientist’s data and compare their values to mine. So I need hard figures. Now, they say a picture paints a thousand words and so I understand why many papers just visualise their data. But it can be remarkably hard (and also time consuming) to pull numbers from a graph, and even then you’re only creating an estimate. Having the ability to pull up original data sets would be unbelievably useful right now.
Which is why I was interested to read about the latest offering from F1000 – F1000 research (for a brief intro to F1000’s other work see the image at the end of this article). F1000 research is due for launch later this year, and their plan is one of post-publication peer-review. This involve submission of articles which will be date stamped and then available for all to see. Reviewers will be selected and the article reviewed, but the entire review process will be visible online.
This is not entirely novel, journals such as Atmospheric Chemistry and Physics, ACP) include discussion space where reviewers and registered users can comment on papers or read the comments of others, and authors are invited to respond, offer additional information or amendments. What excited me most about F1000 research (which I first read about on the Wellcome Trust blog, here) was their aim to encourage…
“the submission of a range of article formats, including standard articles, posters or slides, and content types, such as negative results or case studies. Furthermore, we will be strongly encouraging our authors to publish accompanying data – either separately, or along with the main article. Datasets can also be published, without any associated analysis and conclusions, following basic protocol templates.”
Last year, at the British Science Association Science Communication Conference there was lots of discussion about the ‘Future of Online’. A major worry amongst scientists is that, whilst they realise the benefits that can come from sharing their findings online, they’re worried about getting the credit and how to secure citations for your work. Sure, we should probably all be worrying less about who’s the data is and more about what it tells us about the world/cancer/climate change but the fact of the matter is your data and your publications are your academic passport and without them it’s hard to succeed. Whether that’s right or wrong is a discussion for another day.
Services such as F1000 research would make it easier not only to share your data, but for potential users to search for the data they want and to cite it when they do. It’s also interesting to see them flag the idea of sharing negative or null results; a common problem in fields such as health, food and pharmacy where negative findings never make it to publication.
Unfortunately the F1000 suite of projects is aimed mainly at biology and medical research and so will not be useful across the science board. But if projects like this are a success it is likely that similar projects will start in other research areas. There’s also the issue of payment, costs are likely to be attached to article submissions to F1000 research. Will researchers, faculties and universities be happy to fund publications of datasets or null results? Will publication of a dataset or preliminary result preclude later publication of a full article in a more traditional article? To their credit, the F1000 research team are addressing issues such as this through their regularly updated website, so please check it out for more information.
From my own experience, pulling out my old papers I downloaded in 2009, if we want to increase the success of these projects we need to start the scientific literature learning process earlier. During my first degree we were told to use journals where possible, how to speedily search using Web of Knowledge and how to cite articles. But very little was taught in terms of evaluating the impact of journals, looking for supplementary information or alternative information sources that may be available. At the start of my PhD I swapped Web of Knowledge for Scopus on the back of a convincing sales pitch and some free playing cards but, again, the emphasis was on easy, speedy searches and downloads. Scopus allows you to download whole swathes of articles without even reading the abstracts. Certainly it makes life easier, but it does little to flag up supplementary information, data and graphs that may be available.
What do you think? Do you have any experience using alternative publication or information sources (as an information source or an end-user?).
F1000’s suite of services include a poster repository, post-publication reviews of biological and medical research papers and their own journals.
Please note, I have never used any of the F1000 products and am writing this based on information from their websites/other reviews. If anyone wants to share their experience of F1000 please feel free!