Start Date
17-6-2025 9:00 AM
End Date
17-6-2025 10:30 AM
Abstract
Introduction: Autistic people, including the growing population of autistic college students, have reported low levels of physical activity. (Thompson et al., 2023; McLeod et al., 2021) This is despite physical activity bestowing significant benefits to autistic people, as reflected in physical activity being deemed an evidence-based practice for this population. This study aimed to evaluate the feasibility and acceptability of the ‘MoveMuse’, an artificial intelligence (AI) chatbot co-designed with autistic college students to provide a highly individualised and scalable brief physical activity intervention.
Method: Autistic college students aged 18 or older, attending Irish, Polish and Belgian Universities participated in the study. Each participant interacted with ‘MoveMuse’. This chatbot is powered by ChatGPT and trained on scientific literature. Underpinned by the health belief model and fantasy realization theory, ‘MoveMuse’ guided users through an educational conversation about benefits of physical activity and through a mental contrasting with implementation intentions exercise. A mixed-methods approach, including surveys, think-aloud activities, and semi-structured interviews, were utilised to assess the feasibility and acceptability of this intervention modality.
Results: The mean standardised chatbot usability score was 64.45 (SD=10.55), indicating positive perception of usability. Interview findings suggest a preference for customisable message tones and user interface. Participants commonly expressed preference for less constrained chatbot functions during think-aloud activities. The desire for a personalised planning and scheduling chatbot function for achieving goals was commonly expressed after interactions with ‘MoveMuse’.
Conclusions: These findings provide insight into the potential, and challenges, of using chatbots for providing brief health interventions for autistic populations.
Introduction: Autistic people, including the growing population of autistic college students, have reported low levels of physical activity. (Thompson et al., 2023; McLeod et al., 2021) This is despite physical activity bestowing significant benefits to autistic people, as reflected in physical activity being deemed an evidence-based practice for this population.This study aimed to evaluate the feasibility and acceptability of the ‘MoveMuse’, an artificial intelligence (AI) chatbot co-designed with autistic college students to provide a highly individualised and scalable brief physical activity intervention.
Method:Autistic college students aged 18 or older, attending Irish, Polish and Belgian Universities participated in the study. Each participant interacted with ‘MoveMuse’. This chatbot is powered by ChatGPT and trained on scientific literature. Underpinned by thehealth belief model and fantasy realization theory, ‘MoveMuse’ guided users through an educational conversation about benefits of physical activity and through a mental contrasting with implementation intentions exercise.A mixed-methods approach, including surveys, think-aloud activities, and semi-structured interviews, were utilised to assess the feasibility and acceptability of this intervention modality.
Results:The mean standardised chatbot usability score was 64.45 (SD=10.55), indicating positiveperception of usability. Interview findings suggest a preference for customisable message tones and user interface. Participants commonly expressed preference for less constrained chatbot functions during think-aloud activities. The desire for a personalised planning and scheduling chatbot function for achieving goals was commonly expressed after interactions with ‘MoveMuse’.
Conclusions: These findings provide insight into the potential, and challenges, of using chatbots for providing brief health interventions for autistic populations.
Recommended Citation
Rodgers, Gary; Sweeney, Mary Rose; Staines, Anthony; van Biesen, Debbie; Morgulec-Adamowicz, Natalia; Ogonowska-Słodownik, Anna; and Healy, Sean, "Feasibility and Usability Testing of ‘MoveMuse’, an AI Chatbot to deliver a Brief Physical Activity intervention for Autistic College Students" (2025). International Symposium of Adapted Physical Activity and International Symposium on Physical Activity and Visual Impairment and Deafblindness. 10.
https://sword.cit.ie/isapa/2025/day2/10
Feasibility and Usability Testing of ‘MoveMuse’, an AI Chatbot to deliver a Brief Physical Activity intervention for Autistic College Students
Introduction: Autistic people, including the growing population of autistic college students, have reported low levels of physical activity. (Thompson et al., 2023; McLeod et al., 2021) This is despite physical activity bestowing significant benefits to autistic people, as reflected in physical activity being deemed an evidence-based practice for this population. This study aimed to evaluate the feasibility and acceptability of the ‘MoveMuse’, an artificial intelligence (AI) chatbot co-designed with autistic college students to provide a highly individualised and scalable brief physical activity intervention.
Method: Autistic college students aged 18 or older, attending Irish, Polish and Belgian Universities participated in the study. Each participant interacted with ‘MoveMuse’. This chatbot is powered by ChatGPT and trained on scientific literature. Underpinned by the health belief model and fantasy realization theory, ‘MoveMuse’ guided users through an educational conversation about benefits of physical activity and through a mental contrasting with implementation intentions exercise. A mixed-methods approach, including surveys, think-aloud activities, and semi-structured interviews, were utilised to assess the feasibility and acceptability of this intervention modality.
Results: The mean standardised chatbot usability score was 64.45 (SD=10.55), indicating positive perception of usability. Interview findings suggest a preference for customisable message tones and user interface. Participants commonly expressed preference for less constrained chatbot functions during think-aloud activities. The desire for a personalised planning and scheduling chatbot function for achieving goals was commonly expressed after interactions with ‘MoveMuse’.
Conclusions: These findings provide insight into the potential, and challenges, of using chatbots for providing brief health interventions for autistic populations.
Introduction: Autistic people, including the growing population of autistic college students, have reported low levels of physical activity. (Thompson et al., 2023; McLeod et al., 2021) This is despite physical activity bestowing significant benefits to autistic people, as reflected in physical activity being deemed an evidence-based practice for this population.This study aimed to evaluate the feasibility and acceptability of the ‘MoveMuse’, an artificial intelligence (AI) chatbot co-designed with autistic college students to provide a highly individualised and scalable brief physical activity intervention.
Method:Autistic college students aged 18 or older, attending Irish, Polish and Belgian Universities participated in the study. Each participant interacted with ‘MoveMuse’. This chatbot is powered by ChatGPT and trained on scientific literature. Underpinned by thehealth belief model and fantasy realization theory, ‘MoveMuse’ guided users through an educational conversation about benefits of physical activity and through a mental contrasting with implementation intentions exercise.A mixed-methods approach, including surveys, think-aloud activities, and semi-structured interviews, were utilised to assess the feasibility and acceptability of this intervention modality.
Results:The mean standardised chatbot usability score was 64.45 (SD=10.55), indicating positiveperception of usability. Interview findings suggest a preference for customisable message tones and user interface. Participants commonly expressed preference for less constrained chatbot functions during think-aloud activities. The desire for a personalised planning and scheduling chatbot function for achieving goals was commonly expressed after interactions with ‘MoveMuse’.
Conclusions: These findings provide insight into the potential, and challenges, of using chatbots for providing brief health interventions for autistic populations.
