Using Altmetric to demonstrate the success of Open Access (OA) Books in the Humanities and Social Sciences

Mike Taylor
Head of Data Insights, Digital Science/Altmetric & PhD Student with Professor Thelwall, University of Wolverhampton
a bar chart

Over the last few years, one of the leading criticisms of the OA movement is that it seems to prioritise research in the so-called STEM fields (Science, Technology, Engineering and Medicine) at the expense of the Arts, Humanities and Social Sciences. To some extent, this is a valid criticism, although it can be partially justified by understanding that STEM research is by far, the larger proportion of research outputs, particularly in English. Where we have seen research into the benefits of OA publishing – in terms of increased downloads, citations – or broader and social impact (as measured by altmetrics) – has largely been focused publications in the so-called STEM research areas, and more generally, on publications appearing in academic journals. Kim Holmberg at the University of Turku, Finland published an excellent article on some of the nuanced effects between OA and non-OA journal papers last year (link to Kim’s paper below).

The Arts, Humanities and Social Sciences have a very different profile from that of the STEM fields. The languages and media are more diverse, their outputs have impact over a longer period of time, citations are slower to appear, and researchers in these fields often prefer to publish research in book form. 

What effects the propagation of research within society? 

As part of my research programme with Professor Michael Thelwall of the University of Wolverhampton, I’ve been looking at how research findings are propagated through society, what variables are important to the distribution and diffusion of research findings, and how we can use Altmetric to understand them. 

A key variable to help us understand how propagation happens is OA status, and last month, I published a paper in Scientometrics journal that shows that OA books and chapters get significantly higher levels of attention than their non-OA equivalents, across a broad range of internet attention (including news, blogs, Twitter and policy documents). 

This information may prove to be important for two reasons: 

  1. It can act as a motivation to justify investing in OA books – as it will increase social use, and we can measure this effect. 
  2. We can think about how we can apply research into altmetrics as a way of measuring the effectiveness of our engagement strategies. 
a bar chart
The proportion of books in Arts, Social Sciences and Humanities with attention on Twitter, comparing OA versus non-OA books.

How can we practically make use of Altmetric to understand and demonstrate the successes of OA books, chapters and journals? 

First, we need to recognize the insights that the research by Kim Holmberg and myself has shown. Secondly, you may be looking to compare publisher against publisher – from the perspective of a funder, a publisher or a research institution. Fortunately, Altmetric Explorer’s search and filtering tools make this kind of analysis a snap. My favourite approach is to download the data and use Excel, but of course, Altmetric Explorer will allow you to download data into many business analytical tools. 

Although there is a definite OA advantage for many disciplines and attention sources, it’s not universal. So when you’re reflecting on (for example) the proportion of your articles that get News coverage or the average of Twitter mentions, you need to take into account the field and age, as well as the type of document you’re comparing. Look at what the global population does, and compare it to yours.  

Although the underlying mechanism into the OA Altmetric Advantage is poorly understood (I discuss a few theories at the end of my paper), it does represent a clear advantage and justification for OA publishing. All of us who are involved in the Open Science movement can use this data to express this advantage, and to compare the performance of different journals, publishers and OA policies. As the research shows: this is a complex phenomena, but one that can be successfully shown using Altmetric data. 

The final version can be read here and cited using its DOI 10.1007/s11192-020-03735-8 

This blog post was adapted from one published at the University of Wolverhampton’s website. 

Holmberg, K et al; (2019) Do articles in open access journals have more frequent altmetric activity than articles in subscription‑based journals? An investigation of the research output of Finnish universities? Scientometrics 

Taylor, M; (2020) An Altmetric Attention Advantage for Open Access Books in the Humanities and Social SciencesScientometrics  

Register here to receive the latest news and updates from Altmetric