Monday, June 23, 2014

Impact of open data in archaeobotany?

Thinking about the "impact" of sharing my archaeobotanical data so far. Here are a few statistics as of the evening of 23 June 2014

Figshare statistics

So far the documents uploaded to Figshare (25 of them) have been viewed 430 times.Sometimes this was me checking links. Papers are more likely to be downloaded than datasets (this is a pity, I was hoping that people would do nice stuff with the data). Nobody is sharing the papers and dataset. I have not filed out a profile and used Figshare as a work space, or a networking platform. Most views must be generated by results within Figshare's search facility. I'm not sure what to make of these statistics. One curiosity is the fact that the archaeobotanical report from Ballinglanna North 1 had only 3 views (the lowest of all of my documents on Figshare, and all of those views almost certainly by me), while the dataset for the same site had 34 views (one of the highest hit counts). Odd. I begin to suspect that many of these "views" are not by people... It's hard to know, therefore,whether this data sharing exercise is having an "impact" at all. My Figshare data is available to view here.

Zenodo statistics

I have 12 documents uploaded to Zenodo. In some cases this includes different versions of what is essentially the same report. Zenodo does not currently provide statistics for viewing and downloads, so there are no potential impact measurements at this site. My uploads available to view here.

JOAD statistics

My article written for the Journal of Open Archaeological Data was published in May this year, referencing datasets and reports stored in Zenodo. After approximately six weeks online this has 27 views and 7 downloads. I have not promoted this online, but have sent an email to a few colleagues who may have been interested. Here's the ref.: Johnston, P 2014. Archaeobotanical Data from Two Middle and Later Bronze Age Round House Sites in Cork, Ireland. Journal of Open Archaeology Data 3:e1, DOI:

What's the impact?

It's difficult to assess whether this is having an impact or not. Certainly the datasets don't appear to have been used yet and the journal article hasn't been cited. However, it is likely that measuring this kind of impact is something that really occurs over the long term. Wait and see. Nevertheless, I am curious about how those high viewing figures are generated on Figshare.