Leading the Crowd: The Impact of Transformational Leadership and Emotional Capabilities in Conversational Crowdsourcing

Abstract

Artificial intelligence (AI)-based technologies like large language models (e.g., GPT4) and conversational agents (CA) enrich human work environments and assist workers. In crowdsourcing, CAs can supervise crowdsourcing microtasks to provide a better experience for crowdworkers which is called conversational crowdsourcing. In human relationships, transformational leadership and emotional capabilities by supervisors increase the quality of this unique relationship. Analogous, conversational crowdsourcing may provide social support through leadership and improve the relationship between the leader and the follower, called leader-member exchange (LMX). To analyze this phenomenon, we investigated the effect of conversational crowdsourcing with transformational leadership and emotional capabilities on LMX. We conducted a between-subjects field experiment with 239 participants that performed an idea generation task. Our results show an ambivalence between positive effects on crowdworkers’ LMX, perceived effectiveness, and enjoyment and negative effects on generated idea quality and elaboration. Further, emotional capabilities reduced the stimulating effect of transformational leadership on LMX. With this study, we contribute a theoretical understanding of transformational leadership in conversational crowdsourcing and the phenomenon of its ambivalent effects. We provide a foundation for future research on the application of conversational crowdsourcing and leadership in crowdworkers’ reality.

Keywords: Conversational crowdsourcing, Transformational leadership, Leader-member exchange, Emotions, Chatbot