User reactions to COVID-19 screening chatbots from reputable providers

Area of Science:

• Human-computer interaction
• Public health technology
• Digital health

Background:

• The COVID-19 pandemic necessitated rapid development of remote screening tools.
• Chatbots emerged as a scalable solution for initial patient assessment.
• Understanding user perceptions of chatbot effectiveness is crucial for adoption.

Purpose of the Study:

• To investigate user responses to coronavirus disease 2019 (COVID-19) screening chatbots.
• To identify factors influencing user perceptions of chatbot performance compared to human agents.

Main Methods:

• An online experiment involving 371 participants.
• Participants evaluated COVID-19 screening sessions involving a chatbot or human agent.
• Scenarios included users with mild or severe symptoms.

Main Results:

• Perceptions of agent ability were the primary driver of user response to screening hotlines.
• When ability was perceived as equal, users viewed chatbots neutrally or more favorably than humans.
• Trust in the hotline provider significantly influenced perceived agent ability, with a slight bias against chatbots.
• Asian participants reported higher perceived ability and benevolence than White participants.

Conclusions:

• High-quality service from COVID-19 screening chatbots is necessary but not sufficient for widespread adoption.
• Emphasizing chatbot ability and assuring users of parity with human agents is essential for increasing adoption.
• Addressing potential biases in perception based on user demographics may be important.