This is a guest post by Ros Pyne, Head of Policy and Development, Open Research at Springer Nature. The author would like to thank Mithu Lucraft, Agata Morka and Christina Emery for their contributions to the original report.
Of the hundreds of members of OASPA focussed on disseminating research through open access journal articles, few also publish books. Books are more complex; in most cases they require a greater editorial investment, serve different communities with very different needs, and there is less funding available. But there is an assumption from many that open access will bring the same benefits for books as it does for journal articles, with increased downloads, views and potentially even citations.
Where there are undeniable challenges in introducing a sustainable OA books model, we also see opportunities to drive open access forward and advance discovery through experimentation. As of October 2017, Springer Nature has published more than 400 open access books on SpringerLink, from our SpringerOpen and Palgrave Macmillan imprints. This means that we have a solid and growing dataset from which to investigate the so-called ‘OA effect’.
We recently reviewed a sample of 216 Springer Nature OA books and 17,124 non-OA books included in the download analysis (using SpringerLink data), and 184 OA books and 14,357 non-OA books in the citations and mentions analysis (using data from Bookmetrix).1
The report, launched last week, which came out of this analysis also contains qualitative feedback from authors and funders. It shows that open access books are:
- Downloaded seven times more: On average, there are just under 30,000 chapter downloads per OA book within the first year of publication, which is 7 times more than for the average non-OA book.
- Cited 50% more: Citations are on average 50% higher for OA books than for non-OA books, over a four-year period.
- Mentioned online ten times more: OA books receive an average of 10 times more online mentions than non-OA books, over a three-year period.
The most commonly cited reasons for authors to publish open access included easy access to research; wider dissemination; and ethical motivations. Many considered their subject matter when considering whether or not to publish open access, hoped to secure a cheaper edition of the print book, or thought it would mean a faster publishing time.
Funders had two very clear motivations in funding open access books: either ethical motivations, arguing that publicly funded research should be open for the public to read, or satisfying their authors’ expectations when it comes to reaching a larger audience.
But neither the authors nor the funders we interviewed felt sufficiently informed about the implications and/or benefits of publishing a book via an OA model. They felt they didn’t have the right tools to measure their own investment; this is something that, as publishers, we can work on delivering through platforms such as Bookmetrix.
The positive news is that almost unanimously, the funders we interviewed saw OA as the future of publishing and agree that they would like to see a transition to a system in which all scholarly books are made accessible via an OA model (of course, some of the funders we interviewed are leaders in this field and cannot be considered representative).
We found that there is clear evidence of an OA effect for open access books, and authors and funders appreciate the benefits of OA books, even if the market is not yet mature in the same way as it is for journal articles. But more research needs to be done, and publishers can do more to inform authors and funders about the OA effect, and other implications of publishing books via OA models. If we continue to experiment to find a sustainable model which works, there is an appetite there to build a true and lasting open research ecosystem for both books and journals.
1. We have released aggregated data for all analysis which is available in PDF and excel formats from the report landing page (we are unable to release the full data for title by title analysis, as the non-OA title data is commercially sensitive). It’s worth noting that although the report finds a positive correlation between OA books and higher downloads, it acknowledges that causation cannot conclusively be proved. Open access is a relatively new business model for books, and while have a good dataset, there is at this stage insufficient data to give a complete overview of an OA book’s life. We acknowledge that there are limitations to our initial study and these are discussed further in the report.