Evaluating and designing symptom checkers: a user-centered approach
Open Access
- Author:
- You, Yue
- Graduate Program:
- Information Sciences and Technology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 23, 2024
- Committee Members:
- Jeffrey Bardzell, Program Head/Chair
Xinning Gui, Chair & Dissertation Advisor
Saeed Abdullah, Major Field Member
Chun-hua Tsai, Special Member
Kenneth Huang, Major Field Member
S. Shyam Sundar, Outside Unit & Field Member - Keywords:
- Human-computer Interaction
Health Informatics
Symptom checker
Conversational agent
Explainable AI
Interaction design
Medical authority - Abstract:
- In recent years, there has been increasing interest in developing consumer-facing symptom checker (SC) apps, which can provide potential diagnoses for users by using human-like conversations or a questionnaire that interacts with users. Despite developers of SC apps promising various benefits, such as accurate diagnosis, the use of SC apps may pose risks for users, if lay users blindly trust these diagnoses from SC apps. Therefore, user perceptions and the design of symptom checker apps are crucial, influencing users’ trust-building and further decision-making. However, little research has been undertaken to investigate user experience with these apps in everyday practice and to explore user needs in the interactive design of these apps. To bridge these research gaps, this dissertation investigates the user experience with SCs in real-world settings and examines how their interaction design affects user perceptions by applying a user-centered approach. Through a series of studies, this dissertation pinpoints design opportunities for enhancing SCs and healthcare applications. This work first uses a systematic review to reveal current research gaps and set the direction for subsequent studies. Subsequent interviews with SC users delve into their experiences and needs, elucidating how users assess SCs based on perceived authority of developers, interaction design, and comparisons with traditional healthcare providers. The findings also highlight a user preference for human-like features, explanations, and efficiency. These findings lead to two controlled experiments to further explore design solutions to meet users’ requirements. Overall, this dissertation enhances our comprehension of user needs and factors affecting consumer-facing technology services, offering significant design implications for their future development. Contributing to the domains of Human-Computer Interaction and Health Informatics, this work advances our knowledge of consumer interactions with healthcare technology and delineates the challenges in chatbot design.