Designing for explanation-driven trust in Chatbots

Designing for explanation-driven trust in Chatbots

Zhang, Zhengquan (TU Delft Industrial Design Engineering)

Schneegass, C. (mentor)
Tsiakas, K. (mentor)

Delft University of Technology

Delft University of Technology

Strategic Product Design

2023-08-15

Trust plays an important role in the implementation of chatbot technology. This study was also focusing on user trust in chatbots, particularly focusing on the role of response delay and explanation-driven driven subjective transparency. This research includes a pretest and a main test. In the pretest, we selected one explanation that was perceived by the participants that can raise the most social presence feeling as well as the subjective transparency of the chatbot. In the main test, a 2 × 2 between-subject experiment was designed and conducted to test the hypotheses. First, the findings revealed that while response delay did not significantly influence trust or social presence, clear explanations, especially in the context of instant delays, positively impacted subjective transparency and trust. Second, the study reinforced the positive correlation between social presence and trust, subjective transparency, and trust. From a practical perspective, the research offers insights for chatbot design, emphasizing the importance of improving subjective transparency and rendering a more natural and human-like interaction.

Artificial intelligence (AI)
Generative AI
Response delay
Social presence
Subjective transparency
Chatbot trustworthiness
User Trust

http://resolver.tudelft.nl/uuid:99884f56-3956-4f0d-9e95-95168f6716ba

Student theses

master thesis

© 2023 Zhengquan Zhang