The last de ade has seen two signif ant phenomena emerge in resear h ommuni ation: the rise of open a ess (OA) publishing, and eviden e of online sharing in the form of altmetri s. There has been limited examination of the efe t of OA on online sharing for journal arti les, and little for books. This paper examines the altmetri s of a set of 32,222 books (of whi h 5% are OA) and a set of 220,527 hapters (of whi h 7% are OA) indexed by the s holarly database Dimensions in the So ial S ien es and Humanities. Both OA books and hapters have signif antly higher use on so ial networks, higher overage in the mass media and blogs, and eviden e of higher rates of so ial impa t in poli y do uments. OA hapters have higher rates of overage on Wikipedia than their non-OA equivalents, and are more likely to be shared on Mendeley. Even within the Humanities and So ial S ien es, dis iplinary diferen es in altmetri a tivity are evident. The efe t is onfrmed for hapters, although sampling issues prevent the strong on lusion that OA fa ilitates extra attention at whole book level, the apparent OA altmetri s advantage suggests that the move towards OA is in reasing so ial sharing and broader impa t.
Introdu tionTwo of t e largest p enomena in scientific communication in t e last decade ave been t e rise of Open Access (OA) journal publications, and t e researc area known as altmetrics. OA publications are t ose t at are freely available on t e internet, via a range of routes. Altmetrics is t e collection, reporting and analysis of attention being paid to researc publications across a variety of online platforms.Books and book c apters are under-represented in t e growing corpus of researc on OA. It as been suggested t at t is as arisen due to t e general lack of attention paid to bot t e book form, and also t e Arts and Humanities disciplines -w ic tend to favour books as t eir preferred c annel for publis ing researc .In general, t e absence of sales figures, reliable metadata, download figures and t e relatively slow citation performance of books as made a comparative study of OA versus non-OA books and c apters c allenging. T e relative paucity of data may ave contributed to a low level of examination in t e scientometric literature, and as suc , t ere is a lack of compelling evidence to drive t e adoption of OA for books and c apters.An emerging literature examining t e effect on social s aring and broader impact on OA journal articles offers some met odological insig ts, but given t e known differences