| This dissertation studies the extent to which certain cognitive processes mediate the influence of social structure in determining entrepreneurial behavior by academic scientists. Entrepreneurial behavior is defined as the pursuit and exploitation of opportunities to create value. A network-cognitive model of collaborative entrepreneurship is proposed that conceptualizes social structure as the individual's network of advice and collaborative relationships. Cognitions are modeled as the cognitive biases of illusion of control, representativeness, and overconfidence. The form of entrepreneurial behavior studied is the pursuit and acquisition of external grant funding. The model is estimated on data obtained from an online survey of 3,489 academic scientists in Research I Universities in six fields of science and engineering that resulted in 1,764 usable responses.;Findings reveal that the some aspects of social structure may influence entrepreneurial behavior because they change the level of individual's cognitive bias. Specifically, the influence of the number of advisors, level of trust in advice relationships and the number of grant writing collaborators on entrepreneurial behavior were mediated by illusion of control, representativeness, and overconfidence bias. These mediated effects were found to decrease the number of grant proposals submitted and to increase the percentage of these grant submissions that received funding awards.;These findings have several implications for theory building. First, entrepreneurial behavior in the context of academic science is more complex and multi-staged than is currently understood. This dissertation's re-conceptualization of entrepreneurial behavior as opportunity seeking may be useful in the study of other manifestations of this behavior in the academic setting than grant seeking behavior. Second, the individual's cognitive biases appear to mediate the extent to which social context determines entrepreneurial behavior. However, the network-cognitive model given here includes a small set of possible cognitive processes that could be involved in this mediation. More research is needed to fully understand the mediative role of cognitions between social context and individual behavior. |
