Cooperation and intra-organisational knowledge sharing are vital for effective organisational functioning. However, inter-work unit competition, work unit subcultures and boundaries frequently inhibit knowledge exchange. There is a need for a greater understanding of the socio-psychological determinants of knowledge sharing. To this end, this thesis examined inter-group and inter-personal determinants of knowledge sharing, and investigated whether social identity-based categorisation strategies could be applied to promote inter-work unit knowledge sharing.
A combination of research methodologies were utilised across the four studies in this thesis. Study 1 utilised in-depth interviews with 17 public sector employees to identify socio-psychological determinants of knowledge sharing. While several issues were identified, the two most prevalent themes were interpersonal and intergroup factors, highlighting the central role social
exchange and social identity processes play in intra-organisational knowledge sharing.
Studies 2 and 3 employed a combination of vignettes and a categorisation manipulation to investigate the effects of social exchange processes (i.e., reciprocity) and identity salience on interpersonal knowledge sharing within and between work units. Study 2 examined the effects of exchange history and group membership (in-out group) on knowledge sharing intentions with a sample of 90 public sector employees. Results showed that employees were more likely to share with co-workers with whom they had a positive exchange history and they were more likely to share with members from their own work unit than with members from other work units. Additionally, on the basis of the common ingroup identity model. Study 2 examined whether invoking a superordinate organisational identity would increase knowledge sharing across work units by overcoming the negative effect of in-group bias
(i.e., silo behaviour). Participants resisted attempts to create a superordinate organisational identity at the expense of a subordinate work unit identity. Indeed, a rebound effect was observed, with employees in the superordinate organisational identity condition identifying more strongly with their work unit identities for distinctiveness.
Study 3, on the basis of the dual identity model, examined whether a categorisation intervention, that simultaneously recognised both employees' work unit and organisational identities could overcome the rebound effect observed in Study 2. Ninety public sector employees were randomly allocated to one of three categorisation conditions (subordinate work unit identity, superordinate organisational identity and dual work unit and organisational identity). In accordance with the dual identity model, I found that redefinition of group membership (i.e., recategorisation) by acknowledging subordinate work unit identities within
a superordinate organisational identity was more effective in facilitating inter-work unit knowledge sharing.
Study 4 replicated and extended the findings of Studies 1, 2, and 3 by using a survey methodology (N=271) to examine the effect of social exchange processes, social identity processes and their interaction on intra-work unit and inter-work unit knowledge sharing. Levels of trust, exchange history and knowledge sharing were all higher within work units than across work units. Support for the role of social exchange processes was provided by the finding that trust mediated the association between exchange history and knowledge sharing. Additionally, support for the interrelationship between social exchange and social identity processes was provided by the presence of several mediation and moderation effects.
This thesis makes several contributions to our understanding of knowledge sharing in organisations. First, it integrates
research from the knowledge management. social exchange and social identity literatures to shed light on the process of knowledge sharing. Second, it demonstrates the importance of considering the intergroup context of knowledge sharing by reconceptualizing silo-behaviour as an intergroup problem. Third, it extends research on knowledge sharing as a social exchange by demonstrating the effects exchange history and trust have on knowledge sharing using different methodologies. Fourth, this thesis extended social categorisation research by testing categorisation strategies with two organisational samples. In doing so, to the best of my knowledge. Studies 2 and 3 are the first to directly manipulate employees' work unit and organisational identity salience. Previous research on recategorisation, while generalized to intergroup relations within organisations, has relied on ad hoc or minimal groups, with research on naturalistic groups limited to student samples. Fifth, this thesis
showed that work unit and organisational identification affect employees' perceptions of social exchanges.
Overall, the four studies provide cumulative evidence that social exchange processes, social identity processes and their interaction play an important role in intra-organisational knowledge sharing. Theoretical implications in reference to social identity research in organisations and practical implications for facilitating intra-organisational knowledge sharing are discussed.