Search

The aim of this thesis is to contribute to the understanding of contextual features of national databases of research output and the role these features play in the understanding of research within the social sciences and humanities. This mixed-methods thesis consists of four separate studies, each one contributing to the understanding of a particular aspect of databases for research output. The focus of the first study is on the identification and description of national databases for research output for social sciences and humanities currently operational in Europe. It is based on data collected through two surveys. The first survey gathered responses from 39 countries in Europe and identified more than 20 national databases. The key finding of the second survey is the identification of a variation in terms of different aspects of comprehensiveness (e.g., research output types, authors, institutions). The second study makes use of bibliometric research methods and explores how the choice of disciplinary classifications for journals in the social sciences and humanities influences bibliometric analysis. The findings of this study show that the choice of disciplinary classifications does not equally affect all analyses. While the changes in the absolute number of publications can be substantial, the changes in the relative number of publications are typically small. The third study also employs bibliometric methods and demonstrates an approach to construct complexity-sensitive book metrics for scholarly monographs. This approach takes into account that scholarly monographs can be complex works consisting of different translations, editions, and physical manifestations and results in a typology of different scholarly monographs: Globally-visible single-expression works, Globally-visible multi-expression works (GVME), Globally invisible works, and Miscellaneous. The fourth empirical study employs qualitative research methods to explore the conceptualisations that underpin national databases for research output. The findings show that there are two different database logics—Enlightenment and New Public Management—that shape the way these databases are designed and run. The main conclusion that can be drawn from this thesis is that contextual features of databases for research output are of utmost importance. Databases for research output are created differently and for different purposes thus setting limits to the uses of these databases. Furthermore, this study highlights that the understanding of national databases for research output requires understanding of both the debates on such databases that take place in different knowledge domains and the specifics of research and research governance where the database for research output is situated.

Typically, publications in the social sciences and humanities (SSH) tend to be written more in national languages as opposed to English or other lingua franca. Similarly, the types of publications often are much more diverse. These two are assumptions that can be best studied using national data and, preferably, by means of comparisons of two or more countries. However, cross-country comparisons using national data, as we will show, contain a range of methodological challenges.Interest in SSH publication patterns led to an exploratory study, the preliminary results of which are presented here. The study aimed to acquire a general insight into differences and similarities in publication patterns within SSH across 13 European countries (among them: Denmark, Finland, Norway, and Sweden). The aim was pursued by means of a questionnaire containing questions on national databases in the respective countries and series of bibliometric indicators (e.g., the share of publications in English, the share of co-authored publications).The preliminary findings point to considerable differences in designs of national databases and practices of data registration and processing. Hence, an accurate interpretation of cross-country comparison of indicators is not achievable without an examination of the role that the differences in, say, classifications of publication types, play in the resulting indicators. Such methodological challenges are the focus of the presentation.

This document provides an overview of European databases and repositories for research output within the social sciences and humanities. This overview is a result of a survey conducted within the framework of European Network for Research Evaluation in the Social Sciences and Humanities (ENRESSH, www.enressh.eu). This overview can be used to identify data sources and/or possible collaborators for bibliometric studies focused on social sciences and humanities or other related topics. Similarly, the insights reported here can be used as a starting point for more detailed exploration of designs and contents of databases for research output within the social sciences and humanities. Finally, the findings can serve as a source of background information for discussions on research evaluation and/or research funding allocation suitable for the social sciences and humanities.

In this study I explored educational research publication patterns in relation to the use of the so-called Norwegian model (NM) at the Faculty of Education in the University of Gothenburg, Sweden. The study included 4090 publications. The approach to analysis was designed to enable a comparison of publication patterns before (2005-2009) and after (2010-2014) the introduction of NM.The findings suggest that the numbers for Level 1 are increasing with a greater rate than those publications that are excluded from the model (see Figure 1). However, such a trend can be identified already before the introduction of the model, thus raising further question on how to interpret this pattern. Further exploration suggests that the annual number of Level 2 publications fluctuate around 25 (SD 6.89) throughout the whole period with the exception for the last year explored (2014) when the annual number increases to 62. A comparison on basis of annual average growth rates show that increase in annual average growth rate can be identified only for Level 1 publications. For Level 2 and those publications that are excluded from the model the annual average growth rate after the introduction of the model matches the rate before.Consequently, it is not clear to what extent the change in publication patterns is related to NM. With findings from this, I wish to express a few conceptual and methodological concerns and open up a discussion on alternative ways to capture the relationship between publication patterns and one or another research assessment approach.