Last October I attended the 2:AM Altmetrics conference in Amsterdam. It was a fascinating day (unfortunately I could only attend on the Wednesday), where I met a number of interesting people and gained plenty of inspiration, some of which will hopefully find its way onto The Connected Leiden Researcher in the coming months. This is the first of those posts.
One of the things that often comes to the fore in conversations around altmetrics is that people object to them being used as an evaluation tool. The arguments against are varied, ranging from the fact that they are seen as popularity metrics, as easily gamed, and they currently lack a well-defined set of standards. However, altmetrics can be used for purposes other than evaluation of impact. One of the most important things to realise about altmetrics is that they don't actually measure impact, they measure attention, and it is exactly this latter element that makes altmetrics an interesting discovery tool.
Discover what? And who?
But what exactly can you discover using altmetrics? And what are the questions that you could investigate as a result? To answer this I'd like to take a closer look at the 2014 article by Lin et al. titled Rising Tides or Rising Stars?: Dynamics of Shared Attention on Twitter during Media Events and published in Plos One.
Using a service such as PlumX you can find out where your article was accessed--whether at the journal or from your institution's repository. In the case of an open access journal such as Plos One this might not be relevant, but if you are looking at green open access, where you archive a version of your paper in your institution's repository (with or without an embargo period set by the publisher,) it might give you valuable information about what the effect of making your work available in open access is on its discoverability that could help you decide on where to publish in the future. Additionally, PlumX and Plos both allow you to see how many times people looked at your article online and how many of them clicked through to the PDF (and presumably saved it.) Those numbers can also give you a preliminary idea of the usage of your work, a much earlier one than the traditional citation count.
Someone once told me that it doesn't matter how many people read (or cite) you, as long as the right people read (or cite) you, meaning that for your work to actually have an impact, it has to fall on fertile ground. But who do you want to reach and are you actually reaching them? Altmetric.com and Mendeley can give you a lot of input to answer this question: altmetric.com can show you where the people who tweet about your work are located and whether they are members of the public, scientists, science communicators, or practitioners (based in information they themselves have put in their Twitter bio's.) Similarly, for Mendeley readers they can show you location, profession, and discipline. If you look at the article on Mendeley itself, they show you in which career stage your readers are. Looking at who has been discussing your work on Twitter or Facebook can give you a better idea of whether you are reaching your peers or the general public and whether it has been picked up by special interest or advocacy groups. All three services - Altmetric, PlumX, and Plos ALMs - can show you where your work has been shared.
Looking at the other information provided, what sort of questions can it help you investigate? If you look at who is talking about your work on a more micro level, for example by drilling down into the actual tweets about your work, that can show you who is working on similar topics and perhaps allow you to form new connections that might lead to future collaborations. Or looking at the geographic distribution of your audience, you could explore which territories seem to be accessing your work where you or your research group aren't active yet and find new partners for collaboration. If someone shared your article on Twitter or bookmarked it on Mendeley, what else are they reading and is there material there you hadn't discovered yet? And that is probably just the low-hanging fruit.
Of course the above is all based on the assumption that you look at the metrics for your own work. Indeed, for some services such as PlumX that is the only option, since you don't have access to other people's information. However, in the case of altmetric.com, Mendeley and Plos, you can also look at the information provided for other author's papers, which allows you to look at who is reading them and exploring even more avenues of information, connection, and possibly collaboration. So don't let your view of altmetrics be limited to evaluation; try exploring the possibilities of discovery it offers. You can read more about using altmetric.com's data for discovery on their blog.