Applying Bayesian networks to performance forecast of innovation projects: A case study of transformational leadership influence in organizations oriented by projects

• Published in: Expert systems with applications Vol. 39; no. 5; pp. 5061 - 5070
• Main Authors: de Oliveira, Marco A., Possamai, Osmar, Dalla Valentina, Luiz V.O., Flesch, Carlos A.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.04.2012
• Subjects: Bayesian analysis; Bayesian methods; Continuous improvement; Decision making; Expert systems; Flexibility; Leadership; Networks; New products; Organizational behavior; Simulation; Stimulation; Uncertainty; Organizational behavior; Uncertainty; Bayesian methods; Simulation; New products; Decision making
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2011.11.033

► Predictability of project results is a driver of organizational performance. ► Leadership affects organizational factors that impact performance. ► Bayesian networks are used as a prediction and sensitivity analysis tool. ► Main outcome is the characterization of leadership influence over project performance. The focus of this work is the analysis of the influence of transformational leadership on organizational factors, and their impacts on the project performance. The factors considered are communication, flexibility, continuous delivery and continuous improvement, overlap of activities, and maturity of the team, in projects with a high degree of innovation. Bayesian networks were chosen as a simulation tool. Results showed that for a moderate level of overlap of activities, the maximum project performance is obtained when the leadership components individual consideration, inspirational motivation, idealized influence and intellectual stimulation, are either at moderate levels. This leads to high levels of team maturity, flexibility and continuous delivery, while continuous improvement and communication tend to be moderate. It is highlighted the characterization of the individual contribution of the variables to the project performance and the empirical application of Bayesian networks, as an alternative to statistical methods commonly employed in leadership and management studies.