How Speech Prosody Affects Leadership Perceptions: A Machine Learning Approach

This paper investigates the impact of speech prosody on perceived transformational leadership within a business context. Previous works have mainly concentrated on political preference or leadership; however, the business scenario is rather limited. Notably, a few studies have pointed to the subscale of transformational leadership, yet there is scant research to examine the full-dimensions of transformational leadership. Utilizing a large sample of ۱۲۲ speakers and ۱۲۲ evaluators, OpenSMILE was employed to extract the prosodic features, and Waikato Environment for Knowledge Analysis was used to analyze the data. With the SMO regression as the machine learning algorithm, the results indicated that fundamental frequency, speech intensity, and voicing probability influence various dimensions, including inspirational motivation, idealized influence, intellectual stimulation, and individualized consideration. Fundamental frequency emerges as a pivotal predictor of highlighting the importance of conveying leadership qualities across multiple subscales of transformational leadership. The findings also suggested practical implications for further leadership development by vocal cues. Nevertheless, limitations include the predominantly student sample and the specific context, warranting further research on diverse populations and backgrounds. Future research could explore cross-cultural variations in leadership perceptions and extend analyses to transactional leadership. Overall, this research enriches our understanding of speech prosody in shaping perceptions of leadership and provides insights into the aspects of theory and practice.

Keywords:

Fundamental frequency , Intensity , Voicing probability , Leadership , Machine learning