Measuring the Effect of Mental Health Chatbot Personality on User Engagement

Moilanen, Joonas; Visuri, Aku; Suryanarayana, Sharadhi Alape; Alorwu, Andy; Yatani, Koji; Hosio, Simo

doi:10.1145/3568444.3568464

Cited by 8 publications

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite this, the exploration of trust in live chat encounters has received relatively little scholarly attention. Although trust in conversational agents has been an important area of research [ 43 , 44 ], including traditional communication methods such as telephone conversations [ 45 ], there is still a notable knowledge gap when it comes to complex topics such as cancer. Given the unique complexities that cancer-related questions introduce to the chat interaction, it is essential to understand the specific factors that foster trust in this service.…”

Section: Introductionmentioning

confidence: 99%

Understanding Trust Determinants in a Live Chat Service on Familial Cancer: Qualitative Triangulation Study With Focus Groups and Interviews in Germany

Lanfer¹,

Reifegerste²,

Berg³

et al. 2023

J Med Internet Res

View full text Add to dashboard Cite

Background In dealing with familial cancer risk, seeking web-based health information can be a coping strategy for different stakeholder groups (ie, patients, relatives, and those suspecting an elevated familial cancer risk). In the vast digital landscape marked by a varied quality of web-based information and evolving technologies, trust emerges as a pivotal factor, guiding the process of health information seeking and interacting with digital health services. This trust formation in health information can be conceptualized as context dependent and multidimensional, involving 3 key dimensions: information seeker (trustor), information provider (trustee), and medium or platform (application). Owing to the rapid changes in the digital context, it is critical to understand how seekers form trust in new services, given the interplay among these different dimensions. An example of such a new service is a live chat operated by physicians for the general public with personalized cancer-related information and a focus on familial cancer risk. Objective To gain a comprehensive picture of trust formation in a cancer-related live chat service, this study investigates the 3 dimensions of trust—trustor, trustee, and application—and their respective relevant characteristics based on a model of trust in web-based health information. In addition, the study aims to compare these characteristics across the 3 different stakeholder groups, with the goal to enhance the service’s trustworthiness for each group. Methods This qualitative study triangulated the different perspectives of medical cancer advisers, advisers from cancer support groups, and members of the public in interviews and focus group discussions to explore the 3 dimensions of trust—trustor, trustee, and application—and their determinants for a new live chat service for familial cancer risk to be implemented at the German Cancer Information Service. Results The results indicate that experience with familial cancer risk is the key trustor characteristic to using, and trusting information provided by, the live chat service. The live chat might also be particularly valuable for people from minority groups who have unmet needs from physician-patient interactions. Participants highlighted trustee characteristics such as ability, benevolence, integrity, and humanness (ie, not a chatbot) as pivotal in a trustworthy cancer live chat service. Application-related characteristics, including the reputation of the institution, user-centric design, modern technology, and visual appeal, were also deemed essential. Despite the different backgrounds and sociodemographics of the 3 stakeholder groups, many overlaps were found among the 3 trust dimensions and their respective characteristics. Conclusions Trust in a live chat for cancer information is formed by different dimensions and characteristics of trust. This study underscores the importance of understanding trust formation in digital health services and suggests potential enhancements for effective, trustworthy interactions in live chat services (eg, by providing biographies of the human medical experts to differentiate them from artificial intelligence chatbots).

show abstract

Section: Introductionmentioning

confidence: 99%