Usability Evaluation of Self-Diagnosis User Interfaces for Depression: A Comparative Study among Paper-Based, Text-Based, and Voice-Based Conversational Interfaces

Abstract

Depression is a highly prevalent illness not only in Korea but alsoaround the world. Recently, Artificial Intelligence (A.I.) has been used and contributed in the area of mental healthservice as well. This study aims to identify the most preferred platform for evaluating mental health conditions, particularly depression by analyzing the usability of 3 different platforms; (i)paper-based, (ii)text-based chatbot and (iii)voice-based chatbot. The indicators of usability considered in this study are effectiveness, learnability, reliability, humanness, and likability. Primary data in this paper gathered via survey andface-to-face interview with 50 participants. The result shows significant difference between humanness and likability across all platforms, however, there is no significant difference between text-based chatbot and voice-based chatbot in the aspect of effectiveness, learnability, and reliability. Owing to the highest mean, voice-base chatbot is the most preferred choice of the respondents. Thus, it is predicted that voice-based chatbot will become widely available in the near future as it is effective, easy to learn, reliable, human like and likable. Hence it should be taken into account in the field of depression therapy in the near future.