Sunday, 9 February 2014

Reference manager tools: how well do their support data citations?

Did you miss the last data citers catch-up? If so, we missed you! But here I shall attempt to bring you up to date by summarising what was discussed using this rather neat ‘community blog’. The idea behind the blog is to enable members of the data citers community to contribute a blog post whenever they feel the need – it’s a great idea that I hope it will contribute to a rich discussion around data citation.

If you have never been to a data citers catch-up, you might be wondering what they are like. Imagine a bunch of people from different Australian research institutions who are involved in managing research data and/or building institutional systems to support data management. ANDS set the date and facilitate a virtual catch-up using their gotomeeting software. Usually there is a presentation or two but it also an open forum where people new to data citation can find out more and ask questions from people who will (hopefully) be able to answer them. The questions could be related to anything data citation related – from DOIs to citation element construction to cultural change. The last citers catch-up was a little small, probably because some were attending VALA2014. Here is the gist of our discussion:

If we want researchers to cite data that they use in their research in the same way as they cite publications - leaving aside the argument that data should not be treated in the same way as publications because that’s a whole other topic - then we need to make it part of their research workflows. Reference manager tools assist researchers by proving a means to easily capture and collate citation information for publications (and record links through to those publications). But how do these tools treat datasets? Turns out, most if them don’t or don’t do it particularly well.

Dom Hogan, CSIRO, bedazzled us with a presentation on the heady and (overly?) complex world of reference manager tools and the challenge of type=dataset. Anne Stevenson, also from CSIRO, made an excellent point that the main benefit of incorporating type=dataset into reference manager tools is so that researchers can easily distinguish datasets from publications. I concluded from the presentation that the best reference manager tool to use is Endnote because it does include support for type=dataset. However, to get it work you need to have all of the Endnote patches and updates installed (and it seemed like even then you may have to do some extra steps). 

I think a follow-up to Dom’s presentation would be to create a comparison document for reference manager tools on the basis of their ability to distinguish datasets and capture a data citation. The document could include notes where deemed useful. I also suggest that we (perhaps ANDS on behalf of us) have some direct contact with reference manager tool developers/companies to follow up on whether type=dataset is on their product roadmap and suggest why it should be. I wonder if there is any such document already out there – might be worth checking with our colleagues in UK, Europe and USA. Of course there is a related issue of the need to incorporate data citation guidelines into citation style guides, but this is again a topic for another time.

Problems with reference manager tools aside; plenty of researchers still simply copy and paste a citation from a webpage into a word document [heaves audible *sigh*]. We therefore need to have support for both reference manager tools and display of the citation itself into our institutional data repositories and systems.

Hope to see you at the next data citers catch-up. Check ANDS events for details.


Post a Comment