The impact of social capital on the development of transactive memory systems

This dissertation examines group transactive memory (TM) systems from a network perspective. First, the dissertation develops TM theory from a multilevel perspective. Based on the emergence framework developed by Kozloswski and Klein (2000), the dissertation then reframes TM theory by providing a clear articulation of (a) the basic components and (b) fundamental micro-level processes that lead to the development of TM at the macro level. Second, the dissertation develops TM theory from a network perspective. Building on previous social capital research, the dissertation proposes a multilevel, multi-theoretical framework to study how properties of resource exchange networks shape the development of TM systems at both individual and collective levels. At the individual level of analysis, Burt's (1992) structural hole theory, Granovettor's (1973) strength of weak ties theory, Krackhardt's (1992) strength of strong ties theory, and Lin's (1982; 1995; 2001a; 2001b) social resources theory are utilized to examine how individual social capital shapes the development of individual expertise directories. At the collective level, Coleman's (1988) network closure theory is used to guide the analysis of the impact of collective social capital on the development of transactive memory systems at the macro level. Multilevel data analysis techniques were used to analyze field research data collected as part of a National Science Foundation project.;Results showed that individual social capital had a significant impact on the development of individual expertise directories. Collective social capital and task interdependence did not impact the development of transactive memory systems at the group level, but task interdependence had significant direct impact on collective access to information. It was also found that both network density and task interdependence moderated the relationship between the development of individual expertise directories and individual access to information. It means that in (a) dense networks, or (b) high task interdependence groups, the relationship between the development of individual expertise directories and individual access to information was weaker than in (a) sparse networks or (b) low task dependence groups. Development of TM systems had no moderating effect on the relationships between the development of individual expertise directories and individual access to information.