Paper short abstract:

This paper develops an actor-network theory oriented approach to analyse possible gender patterns with respect to publication rates, studied at the level of departments. A main contribution is in the idea of studying scientific publication as socio-material assemblages.

Paper long abstract:

In STS, gender differences in publication rates of scientists are often linked to marital status, motherhood, women's relative lack of confidence, or discrimination (Fox, 2005; Mason et al., 2013). Others have argued that invisible research specialization is the "missing link" to understanding gender-related publication gaps (Leahey, 2006). However, publication rates differ substantially between departments; therefore, observed gender differences may be a level fallacy. We ask: how does the pattern vary across departments by investigating departmental and related epistemic cultures (Knorr-Cetina 1999) as contexts for analysing publication patterns and gender differences. What do publication rates patterns look like at this level, and to what extent do measures of differences between staff vary between departments? Do we find gender differences also at the departmental level? Furthermore, the paper analyses how patterns of difference are co-produced with other organizational and cultural factors like departmental management, publication policies, and awareness of gender issues. Are there any common features of departments with and without gender differences in publication output? Investigating these questions, the paper challenges common theories of the alleged gender gap. As an alternative, it develops an actor-network theory approach, studying publication patterns as socio-material assemblages (Latour, 2004). The study is based on a database of all scientific publications by person, position, and departmental affiliation of scientific employees at Norwegian University of Science and Technology in the whole period of 2010-2014.