Graduation Year


Document Type




Degree Granting Department

Industrial and Management Systems Engineering

Major Professor

Kingsley A. Reeves


collaborative R&D, innovation networks, network simulation, social network analysis, technological knowledge creation


This research investigates the impact of organizations' collaboration strategies and network positional variables on invention performance. Organizations, particularly those pursuing a differentiation strategy, are motivated to introduce novel products and services in order to remain competitive. Thus, two questions of interest to such organizations regarding the network dynamics of the invention process are: 1) What kind of strategies allow them to attain superior invention results? 2) What is the most advantageous structural positioning in a collaborative network of innovators? Three independent studies attempt to find answers to these questions by using three complementary study approaches.

In the first study, in order to examine invention performance of organizations with different collaboration strategies, a simulation model is constructed and run at various levels of technological dynamism and with various types of invention tasks. The cognitive cooperation strategy, which pursues technological knowledge integration as a priority, is found to be the most effective strategy in most experiments. Success-driven cooperators, whose priority is to collaborate with the most effective performers in the network, provided the most consistent invention performance. Moreover, different strategies are shown to perform best at certain points of the industry environment space spanned by technological dynamism and invention type.

The second study investigates the impact of positional metrics in a collaboration network on the inventive performance of organizations (as measured by the number of patents issued) by using survey data. Twenty-eight high-tech companies and universities located in Florida are surveyed to reveal their collaborative networking map. Network structural measures are obtained in order to test the hypotheses that high values in centrality metrics correspond with higher inventive performance. The regression analyses imply that degree and closeness centralities are predictive indicators of high inventive performance but the data does not support the significance of impact of local clustering.

The third study analyzes the impact of positional metrics on inventive performance by using a national database for the announced research and production joint ventures. From the most inventive organizations (in terms of patent counts) 63 of them are selected for analysis. 371 research and production joint ventures are analyzed to identify their relationship network every year from 1994 to 2012. The results indicate that the number of joint venture partners, being close to other members of the network through network connections and strong local connectivity (except for their interaction effect) is associated with higher invention performance.

All three studies bring new methodological contributions to the area of invention network research. The simulation study offers a new model in the area of collaborative invention networks. Furthermore, the ideas and practices developed during the construction of the agent based simulation model for the invention network can be adapted to similar areas of application. The survey study offers a holistic approach for the definition of connections in the development of invention network map and empirically tests it. The database study longitudinally analyzes the evolution of a highly accurate joint venture network over 19-year period while testing the impact of positional metrics with un-weighted and weighted calculation methods.

Solutions to our health problems, communication or transportation needs etc. are not usually found due to some series of fortunate events. They are the product of an effective recombination process of technological knowledge. Moreover, effective invention performance is not only important for organizations individually, it is crucial for governments that are concerned with the problems of its citizens. Effective ways of facilitating the recombination of technological knowledge are addressed and presented to inform both companies and policy makers. Better understanding of the dynamics of the invention process will bring more solutions to existing problems.