How Does Interactivity Shape Users’ Continuance Intention of Intelligent Voice Assistants? Evidence from SEM and fsQCA

Purpose: With the rapid expansion in the use of intelligent voice assistants (IVAs) in people's daily lives, how to improve users' continuous intention is crucial for the sustainable development of intelligent voice technology. Utilizing the stimulus-organism-response (S-O-R) framework, we propose a theoretical model to examine how three dimensions of interactivity (ie, two-way communication, responsiveness, perceived control) impact individuals' affective reactions (ie, psychological ownership, subjective well-being) and continuance intention of IVAs and how that effect differs technology readiness. Methods: To validate the proposed model, 412 valid samples were collected in China and underwent analysis using a comprehensive approach that incorporated partial least squares-structural equation modeling (PLS-SEM) and fuzzy-set qualitative comparative analysis (fsQCA). Results: The findings from the PLS-SEM analysis indicate that three dimensions of interactivity have significant impacts on affective reactions to varying degrees, thus affecting users' continuance intention. Among these dimensions, responsiveness is the strongest predictor of affective reactions. Additionally, the impact of subjective well-being on continuance intention is stronger when users with high technology readiness. Finally, the results from fsQCA support the PLS-SEM findings and provide three configurations with different combinations of antecedents that sufficiently explain high continuance intention. Conclusion: Our findings reveal the internal mechanisms through which the three dimensions of interactivity impact users' continued usage of IVAs. This study is among the first to examine the effects of dimensions of interactivity on behavioral intentions, utilizing both symmetric (PLS-SEM) and asymmetric (fsQCA) methodologies to identify the most significant dimensions of interactivity and determine sufficient combinations of dimensions to predict users' intention to continue using IVAs. These findings offer valuable and fresh insights for both theoretical understanding and practical application.