Astrom 89 Comparing citation patterns in entrepreneurship research articles in subject handbooks and Web of Science journals Fredrik Åström 1 1 fredrik.astrom@lub.lu.se Lund University Libraries, Head Office, P.O. Box 134, 221 00 Lund (Sweden) Abstract Not knowing enough about the similarities or differences of citation structures between different types of publications creates problems related to whether citation maps of research fields based on Web of Science/ISI data are representative of research fields as a whole or if they are a representation of how WoS perceives the field, not the least in the humanities and the social sciences. To investigate this problem, the citation structures in entrepreneurship research were analyzed using citation data both coming out of WoS-indexed journal articles and citation data from 12 entrepreneurship research handbooks. The datasets were analyzed by studying the age of references, the co-citation structures and also, citation overlaps. The results show substantial similarities between the two data sets: the distribution of the age of the references is almost identical, the co-citation structures in form of co-citation maps have strong similarities both in terms of identifiable networks as well as which cited authors are grouped together and there is also a strong citation overlap between the two different publication types. Introduction Traditionally, visualizations of research areas based different variations of co-citation analyses have been using data from the ISI/Web of Science (WoS) databases, and since the 2000s, also the Scopus database. The consequence of this is co-citation analyses to a large extent building on references coming out of research published in journals, whereas the access to information on research literature published in other forms of publications such as monographs and anthologies has been limited to analyses on WoS non-source items (e.g. Butler & Visser, 2006; Nederhof, van Leeuwen & van Raan, 2010). This means that we are still limited to analyzing references coming out of journal articles. At the same time, we also know that the coverage of the social sciences in WoS is far from complete; and that there are large differences in publication and citation behaviour in different fields of research (e.g. Hicks, 2004; Moed, 2005; van Leeuwen, 2006); and that representations of research fields can vary substantially depending on both selection of material and choice of e.g. level of analysis (Åström, 2002; Åström, 2010; Hellqvist, 2010, Moya-Anegón, Herrero-Solana & Jiménez- Contreras, 2006; Zhao, 2003). When creating maps of research fields in e.g. the social sciences, this creates a problem of lack of control over to what extent the maps based on WoS data are representative of the field we are analyzing as a whole, or if the maps primarily represents subsets of the field publishing research in WoS-indexed journals. Not the least is this question important in terms of the social sciences and the humanities, where a large extent of the research communication is published in other forms of media, such as monographs and anthologies. The aim of this paper is to investigate the citation structures of one field but in two different kinds of publications, approaching the question of to what extent the structures found in co- citation mapping of a field based on WoS data represents the field or just the WoS representation of the field? Within the framework of the Explore project, a database of references from each individual chapter of 12 entrepreneurship research handbooks was constructed (Landström, Hairichi & Åström, 2010), presenting us with an opportunity to compare the citation structures identifiable in these handbooks with analyses made on WoS data.