Abstract
INTRODUCTION: Chatbots can provide task-related services but also act as empathic conversation partners allowing for social interactions. Focusing on such social chatbots, we considered several theoretical frameworks to investigate potential users' intention to use social chatbots and focused on the role of anthropomorphism. METHOD: Based on an online survey with 180 participants, we examined the role of 14 personal characteristics in potential users' acceptance of social chatbots based on bivariate correlations and multiple regression analysis. Based on a subsequent within-subjects experiment and repeated measures analysis of variance, we also investigated differences in the intention of potential users to use more human-like versus less human-like social chatbots regarding their avatar and name. RESULTS: Most personal characteristics were significantly correlated with participants' intention to use social chatbots. The multiple regression model explained about 75% of variance in participants' intention and identified experience and attitude regarding social chatbots as particularly important personal characteristics. Further, perceived usefulness, subjective norm, and perceived behavioral control regarding social chatbots as well as social support showed specifically strong bivariate correlations. The experimental part revealed that more human-like social chatbots received slightly yet significantly higher intention ratings. DISCUSSION: We identified relevant personal characteristics for potential users' intention to use social chatbots and found that potential users prefer using social chatbots with a more human-like appearance. While anthropomorphism can affect potential users' intention to use social chatbots, other aspects seem more important. Overall, our findings provide valuable starting points to better understand why people intend to use social chatbots.