Chatbots supporting mental health and wellbeing of children and young people; applications, acceptability and usability

Research output: Contribution to conferenceAbstractpeer-review

313 Downloads (Pure)

Abstract

Evidence shows that engagement and adherence to mental health chat- bots is typically low. The views of young people around the acceptability and usability of chatbots supporting mental health are important, as mental illnesses typically begin in early adulthood and early intervention is imperative. This study examines the views of young people in relation to four current chatbots (3 symptom management chatbot applications and one positive mental health chatbot); the personality profile that would be desirable for these chatbots and; the situations in which men- tal health chatbots would be beneficial. Participants were mental health service users aged between 16 and 24 years (N=13). Two focus groups were shown demonstrations of four chatbots (ChatPal, Wysa, Youper and Replika). A semi structured schedule and questionnaires were used to elicit data. The chatbots scored on average 3.62 to 5.62 on a 7 point scale to assess whether the chatbot would meet their requirements. Average usability scores ranged from 4.62 to 6.15, based on a 7-point scale. When asked to about the ideal personality for a mental health chatbot (10 point scale). Average scores were openness, 6.75; conscientiousness, 7.5; extraversion, 6.46; agreeableness 8.92; and neuroticism, 1.73. Qualitative data yielded 3 themes; (i) appropriate uses, (ii) desirable features, and (iii) undesirable characteristics. Discussions address the practical challenges associated with meeting the needs and expectations of young people in designing chatbots to support mental health. Examples include the challenge of reflecting desirable personality attributes through text based interventions, balancing the limited capabilities of a text based digital intervention against meeting the individual needs of users, and considering usability in terms of both initial engagement with the chatbot and also long term sustainability. Research directions for this area are proposed.
Original languageEnglish
Pages57
Number of pages1
Publication statusPublished (in print/issue) - 13 Sept 2023
EventEuropean Conference on Mental Health - Ljubljana, Slovakia
Duration: 12 Sept 202315 Sept 2023
Conference number: 11
https://ecmh.eu/

Conference

ConferenceEuropean Conference on Mental Health
Abbreviated titleECMH
Country/TerritorySlovakia
CityLjubljana
Period12/09/2315/09/23
Internet address

Fingerprint

Dive into the research topics of 'Chatbots supporting mental health and wellbeing of children and young people; applications, acceptability and usability'. Together they form a unique fingerprint.

Cite this