Altmetrics are catching a head of steam in science, but are still not widely used.

What are they? Jason Priem, who coined the term “altmetrics”, on almetrics.org defines almetrics as “the creation and study of new metrics based on the Social Web for analyzing, and informing scholarship”.  Find out more here and here and here.

TotalImpact is a project actually providing altmetrics on published papers.  You can query DOIs for papers and they will serve up altmetrics from various sources, including CiteULike, Mendeley, Twitter, blog posts, and more. Definitely go check out what total-impact is up to at their site.

Let’s see what altmetrics on British Ecological Society papers looks like.

I got data from ~ 3,217 papers from the five British Ecological Society journals (Methods in Ecol. and Evol., J. Applied Ecology, J. of Ecology, J. of Animal Ecology, Functional Ecology) using the following protocol:

1) get DOIs for British Ecological Society papers via the Mendeley API in R using the rmendeley package

2) search TotalImpact with each DOI for altmetrics;

3) get some metadata on papers using the CrossRef API (R interface here); and

4) summarize data and plot results.  Get the code at GitHub.

The search was done back in February – apologies for not doing a search more recently, but it took quite a while. What can we learn from altmetrics on BES journals?

The data is quite sparse. Most of the metrics have no data for most DOIs. Nonetheless, I have plotted the data in a few different ways.

The first figure shows mean values of each metric for each BES journal. Error bars are +/- 1 se.

The next figure is the same, but with the Mendeley_readers metric removed to show better the other metrics.

The next figure shows values of each metric over the years of publications for each metric type and each journal (coloured lines).

What are your thoughts on altmetrics?  Good, bad, interesting?  Of course I didn’t include here tweets, facebook likes, etc. – we’ll do that another day.