Hybrid trust in voice assistants

• Applicant: Professor Dr. Marc Ziegele
• Subject Area: Communication Sciences
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 507940485

Voice assistants such as Amazon Alexa and Google Assistant are increasingly being used as digital interaction partners in Germany, for example to search for information or current news. In some cases, users integrate the human-sounding assistants deeply into their daily routines and perceive them as friends or family members. So far, it is not clear how this increasing usage of voice assistants and their integration into daily routines can be explained. The proposed project builds on the concept of trust as a possible explanation. Trust is relevant in many social interactions in which one party expects a performance by another whose actions it cannot fully control. In the proposal, we elaborate on interdisciplinary trust research and develop a new concept of trust ("hybrid trust") that considers voice assistants in their multidimensionality as technology, social actors, and information intermediaries. Based on this concept, the proposed project aims to answer two overarching research questions: 1. How and under which conditions does trust in voice assistants emerge? 2. What role does trust in voice assistants play in their integration into users' daily routines? The first goal of the project is to develop a theoretical model of hybrid trust in voice assistants. For this purpose, we review trust research from social psychology, technology science, and communication science and present a working model that includes determinants, consequences, contexts, and processes of hybrid trust, and allows deriving hypotheses. The second goal is to explore the trust relationships between humans and voice assistants using a method triangulation: First, we use qualitative content analysis to examine online discussions and reviews of voice assistants for evidence that allows inferences about fulfilled and disappointed trust. Second, to further explore trust relationships, we conduct qualitative observations of users in their private environments, followed by interviews. The findings from these studies serve as preparation for the third goal, the quantitative examination of the causes, processes, and consequences of hybrid trust in voice assistants. For this purpose, we present a six-month panel study and product test using both survey data and voice protocols. The findings are used to revise the theoretical model. With the development and comprehensive empirical testing of the new model of hybrid trust, the proposed project transfers existing trust concepts to relationships between humans and digital anthropomorphic interaction partners and thus contributes both to social science trust research and to the understanding of the integration of voice assistants into the everyday routines of users.

• DFG Programme: Research Grants