Koumbarakis, ParisParisKoumbarakis2023-04-132023-04-132021https://www.alexandria.unisg.ch/handle/20.500.14171/111124Scholars agree on the central roles that start-up activities and cognition play with regard to the successful launch of a business. Despite their importance, key questions about how behavioral differences between nascent entrepreneurs impact the outcomes of their ventures remain unanswered. This dissertation attempts to answer some of these fundamental questions by providing empirical evidence of the role that cognition plays in the gestation process and by evaluating machine learning algorithms to predict a venture’s outcome at an early stage. While cognitive and motivational aspects driving start-up activities remain relatively underdeveloped in the current literature, the first paper addresses part of this shortcoming by showing how self-regulation influences the relationship between nascent entrepreneurial exploitation activities, firm birth and firm abandonment. This paper provides evidence of an increase in persistence as well as an increase in the likelihood of achieving firm birth due to a regulatory fit between exploitation activities and a promotion orientation. Next to the conducted start-up activities, the impact of innovation and imitation on a firm’s emergence and survival remains theoretically and empirically inconclusive. The second paper seeks to shed some light on the impact of the rather under-researched imitative business ideas on the likelihood of a firm’s emergence. By drawing on social identity theory, this paper further delineates how founder identity types impact the likelihood of imitating and innovating. The investigations conducted confirm that imitation can be beneficial in an early stage and that certain identity types (i.e., Missionaries) are rather prone to innovate. Finally, this research seeks to predict the likelihood of entrepreneurial success. While few attempts have been made to provide an early stage venture outcome prediction model, the third paper set forth to apply and evaluate a diverse set of machine learning algorithms as well as artificial neural networks to predict the likelihood of firm birth and firm abandonment. The analyses reveal that neural networks can provide a comparable performance in predicting the venture outcome, while especially the XGBoost algorithm provides good performance metrics in predicting the likelihood of a firms’ birth. In sum, the three papers offer valuable contributions to the extant entrepreneurship literature with respect to the impact of cognition on different venture outcomes. By using different machine learning techniques in predicting the venture outcome, this dissertation provides practitioners with a decision support system to optimize corresponding resource allocation.enEntrepreneurshipUnternehmerverhaltenUnternehmensgründungMaschinelles LernenPrognosemodellEDIS-5056doctoral thesis