All entries for Friday 25 January 2013
January 25, 2013
Writing about web page http://altmetrics.org/manifesto/
This time last year, PLoS started to display even more article level metrics. I felt intrigued by the openness about article downloads on PLoS. It led me to wonder whether other publishers were sharing such information so publicly.
Repository tools are available to display accesses for articles in the repository, as at the Bath University repository in this record of a journal article. Some repositories have chosen not to make such statistics publicly visible, however.
I don't see many publishers publicly displaying article level metrics like this, but publishers do sometimes showcase the most downloaded content publicly on their website. For example Springer journal home pages display their most downloaded articles: see the journal Artificial Intelligence
Other publishers share download statistics with the authors, although not publicly displaying them on article records. For example, Nature describe this as a benefit to their authors.
I'd be glad to hear of other examples of publishers displaying these download statistics. I think that authors should be able to monitor activity around their papers for themselves, and I wonder if there is a role for such statistics in helping our readers to ascertain the highest quality papers.
However, I am slightly cautious about download statistics being publicly visible because there are often many versions of a journal article available: on the author's website, on the publisher website and in a repository, for example. And I think that all of these versions should be available because this provides an insurance and archival possibility for authors as well as additional discovery and access options for readers, but that a focus on download statistics could lead authors to become wary of sharing their articles in so many places.
The Internet allows us to create lots of metrics about a researcher or a work, as evidenced by all the altmetrics activity (see my link, above). But just because a measure exists, should it be publicly visible? How should we use these new measures? My own answer is "with caution" and indeed that the author is best placed to make use of the download statistics, because the author will be best able to understand what they mean.