A new tool for fighting the file drawer effect

Until recently, Mark Hahnel was a PhD student in stem cell biology. Frustrated by seeing how much of his own research output didn’t make it to publications, he endeavored to do something about it by developing a scientific file sharing platform called FigShare. Recently, Mark and FigShare were taken under the wing of Digital Science, a Nature Publishing Group spinoff, and a sleek new FigShare was relaunched in January 2012 with many more features and an ambitious scope.

FigShare allows researchers to publish all of their research outputs in seconds in an easily citable, sharable and discoverable manner. All file formats can be published, including videos and datasets that are often demoted to the supplemental materials section in current publishing models. By opening up the peer review process, researchers can easily publish null results, avoiding the file drawer effect and helping to make scientific research more efficient.

Users do not have to pay for access to the content: public data is made available under the terms of a CC0 waiver and other content under CC-BY.  And FigShare is currently providing unlimited public space and 1GB of private storage space for free.

This is a promising solution for getting negative and otherwise unpublished results out into the world (figures, tables, data, etc.) in a way that is discoverable and citable.  Importantly, much of this content would not be appropriate for Dryad, since it is not associated with (and not documented by) an authoritative publication.

There are clearly some challenges to the FigShare model.  A big one, shared with many other Open Science experiments that disseminate prior to peer review, is ensuring that there is adequate documentation for users to assess fitness for reuse.  Another challenge that Dryad is greatly concerned about is guaranteeing that the content will still be usable, and there will be the means to host it, ten or twenty years down the road.  These are reflections of larger unanswered questions about how the research community can best take advantage of the web for scholarly communication, and how to optimize filtering, curating or preserving such communications. To answer these questions, the world of open data needs many more more innovative projects like FigShare.

Considering FigShare’s relaunch suggests a few strengths of the Dryad model:

  • Dryad works with journals to integrate article and data submission, streamlining the deposit process.
  • Dryad curators review files for technical problems before they are released, and ensure that their metadata enables optimal retrieval.
  • Dryad’s scope is focused on data files associated with published articles in the biosciences (plus software scripts and other files important to the article.)
  • Dryad can make data securely available during peer review, at the request of the journal.
  • Dryad is community-led, with priorities and policies shaped by the members of the Dryad Consortium, including scientific societies, publishers, and other stakeholder organizations.
  • Dryad can be accessed programmatically through a sitemap or OAI-PMH interface.
  • Dryad content is searchable and replicated through the DataONE network, and it handshakes with other repositories to coordinate data submission.

For more about Dryad, browse the repository or see Why Should I Choose Dryad for My Data?

A file sharing platform and a data repository are different animals, to be sure; both have a place in a lively open data ecosystem. We wish success to the Digital Science team, and look forward to both working together, and challenging each other, to better meet the needs of the research community.  To see what other options are out there for different disciplines and types of data, DataCite provides an updated list of list of research data repositories.

UK Parliament report supports Dryad and data access

Dryad has won high-level support from the UK Parliament. Its Select Committee on Science and Technology has been reporting on the peer review of scientific publications. Among the questions it considered was:  How far should reviewers be expected to go to assess technical soundness? The report discusses the feasibility of reviewing the underlying data behind research, and how those data should be managed.

Section 4 of the report (para 189) concludes:

If reviewers and editors are to assess whether authors of manuscripts are providing sufficient accompanying data, it is essential that they are given confidential access to relevant data associated with the work during the peer-review process. This can be problematical in the case of the large and complex datasets which are becoming increasingly common. The Dryad project is an initiative seeking to address this. If it proves successful, funding should be sought to expand it to other disciplines. Alternatively, we recommend that funders of research and publishers work together to develop similar repositories for other disciplines.

The Science and Technology Committee concludes that in order to allow others to repeat and build on experiments, researchers should aim for the gold standard of making their data fully disclosed and made publicly available:

Access to data is fundamental if researchers are to reproduce, verify and build on results that are reported in the literature. We welcome the Government’s recognition of the importance of openness and transparency. The presumption must be that, unless there is a strong reason otherwise, data should be fully disclosed and made publicly available. In line with this principle, where possible, data associated with all publicly funded research should be made widely and freely available. Funders of research must coordinate with publishers to ensure that researchers disclose their data in a timely manner. The work of researchers who expend time and effort adding value to their data, to make it usable by others, should be acknowledged as a valuable part of their role. Research funders and publishers should explore how researchers could be encouraged to add this value.

H.M.S.O. Science and Technology  Committee. Eighth Report: Peer review in scientific publications. Published 28 July 2011  Available at: http://www.publications.parliament.uk/pa/cm201012/cmselect/cmsctech/856/85602.htm

Archiving legacy data, or “Why is Dryad better than a floppy disk?”

If you have recently published data in Dryad, chances are it was in the course of publishing an article at a partner journal that steered you our way.

But you may be aware that Dryad accepts data from any peer-reviewed article in biology or biomedicine.  That includes journals that are not (at least not yet) partners.  In fact, as of the the time of writing, Dryad has data associated with articles in 79 journals, approximately four times the number of partners.

Dryad even accepts data from articles that have already been published.  Now, why might you wish to go to the trouble of rummaging through those old files and putting your legacy data online?

Well, we noticed a while back that some individuals were beginning to do this systematically.  For example, there was a sudden influx of data packages with Frédéric Delsuc’s name on them a little while back.  Delsuc, of the French National Centre for Scientific Research (CNRS) and the Université Montpellier, is a member of an international team of collaborators (from France, Norway, Canada, Spain, Japan, Germany, Switzerland, and the United States) that has been using DNA sequence data to reconstruct the evolutionary history of a wide range of vertebrates and vertebrate relatives, from anteaters to sea squirts.

Giant Anteaters

Giant Anteaters (Myrmecophaga tridactyla). The pup clinging to his mother is Cyrano, who was born at the Smithsonian’s National Zoo in 2009. Photo credit: Mehgan Murphy, CC-BY-NC-ND, http://creativecommons.org/licenses/by-nc-nd/2.0/

So far, Delsuc and his team [1] have deposited data from 20 articles in Dryad. The articles are in partner journals such as Molecular Biology and Evolution, Molecular Phylogenetics and Evolution, Systematic Biology, as well as more general science journals such as Nature, Science, and the Proceedings of the National Academy of Sciences USA.

The articles stretch back to 2002, a time when most new desktop computers were still being outfitted with floppy drives. (Remember those?)

We asked Delsuc what he saw as the advantages to archiving his team’s heritage of legacy data?

We […] decided in our team to try to systematically submit our datasets to Dryad because we really think they are valuable. Dryad offers a very nice way of archiving the data ensuring their durability over time.

For Delsuc and his team, no more rummaging through old storage devices to find the files when they receive an email request.  No more worrying about the data when  lab or departmental websites move.  They just need to point their colleagues to Dryad.

It has been reported that the number one reason cited when scientists are asked why they have denied their colleagues’ requests for data in the past was the amount of effort required to dig them up [2].  Delsuc’s and his team intuitively understood that, and went back to archive their data before memories faded, storage devices failed, and graduate students moved on.

The downside to archiving legacy data in this way is that an article’s readers won’t immediately know about the existence of the Dryad data package, since the data DOI will not be published within the text. So, while archiving legacy data has its advantages, there is no substitute for depositing the data before the article is published, as Dryad does with the new articles appearing in its partner journals.

To give Delsuc the final word:

It would be great if more and more journals in the field decide to include data deposit in their publication policies.

[1] Equipe Phylogénie et Evolution Moléculaire” (Phylogeny and Molecular Evolution team) of the Institut des Sciences de l’Evolution (Institute of Evolutionary Sciences), part of the CNRS: Centre National de la Recherche Scientifique (French National Centre for Scientific Research) and the Université Montpellier 2 (University of Montpellier 2).

[2] Campbell EG et al. (2002) Data Withholding in Academic Genetics: Evidence From a National Survey. JAMA 287(4):473-480. doi:10.1001/jama.287.4.473

Why don’t we share data?

There are lots of opinions and answers to this question.  For starters, here’s a lively blog post, responding to this editorial last April.  Consider also this blog post.

What do you think are the barriers to data sharing?

Data from: Thompson S, Daniels K. 2010. A porous convection model for small-scale grass patterns. American Naturalist 175: E10-E15. Dryad Digital Repository. http://hdl.handle.net/10255/dryad.857