Share Your Data!

June 24, 2012

I first heard about Figshare on Twitter earlier this year, but it took me a while to start sharing. The delay was caused by my indecision over which unpublished data to release from the voluminous archive generated over the last five years by all the projects pursued in my lab at Princeton. For the first tranche of uploads, I decided to focus on two types of research output :

(1) supplementary data, e.g., high-resolution electron microscopy images, associated with my lab’s published papers, primarily our recent PLoS ONE paper on membrane accumulation of the SSRI antidepressant sertraline/Zoloft® in yeast cells;

(2) unpublished, icebreaker data intended to be a catalyst for online scientific discussions and new collaborations.

One of the cool features of Figshare is the altmetrics data, specifically the number of times a research object has been viewed or shared, as shown here in a screenshot of a representative single post page (above).

As of this writing, the 19 research objects I’ve uploaded have amassed 3,425 total views and 101 total shares. 70% of shares are tweets, with the remainder split between Facebook and G+. I was particularly curious about the relationship between views and shares, which I graphed here:

Two things struck me. First, nine objects received zero shares. None of them individually garnered more than 60 views, suggesting that sharing is required for large viewership, at least for my data. Second, there is a nice positive relationship between sharing and viewing.

A few qualitative observations might be useful for interpreting this static snapshot. Viewership wanes quickly, with most of the buzz lasting only a few days. However, research objects generously shared continue to amass views at a non-negligible rate compared to the zero-share class, though there is high variance in growth trajectories.

All in all, I’m impressed with the usability of Figshare, and I look forward to its wider adoption. It’s the principal reason why I leveraged the Figshare API on my Open Science platform. To boot, Mark Hahnel, the founder of Figshare, is helpful and timely with technical site assistance, and very generous with retweets!

Related Posts

  • http://abdallahalhakim.tumblr.com/ Abdallah Al-Hakim

    I think FigShare looks like an amazing tool that is addressing an important topic of how scientist can get credit for unpublished or negative data. Also, importantly it allows scientist to reveal the vast amount of data they have which never makes it to publication. This could be the beginning (at least I hope) of a new approach to scientific publishing. The difficult step would be to get universities and funding agencies to recognize these type of applications and crediting the scientists who share their data.