Date of Award
5-2025
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Psychology
Committee Chair/Advisor
Dr. Kaileigh Byrne
Committee Member
Dr. Dawn Sarno
Committee Member
Dr. Sabarish Babu
Abstract
Rates of depression in college students have skyrocketed since the beginning of the COVID-19 pandemic, reflecting a need for effective, accessible mental health interventions. One potential intervention includes mental health (MHealth) apps. Virtual agents within MHealth apps may represent one way to provide human-like connection and therapeutic benefits. Despite the abundance of MHealth apps currently available for public use, few include virtual agents. Moreover, it is unclear which virtual agent features within mental health apps may be particularly beneficial in providing a positive user experience. The current study sought to evaluate how features of virtual agents, such as animation and conversation, impact the usability and user experience of a brief cognitive behavioral therapy (CBT)-based MHealth app. The study also aimed to explore whether differences in these effects exist between users who are experiencing depressive symptoms and users who are not. College student participants were randomized to one of four conditions: conversational and animated virtual agent, conversational and non-animated virtual agent, animated and non-conversational virtual agent, or non-conversational and non-animated virtual agent. Over the course of two weeks, participants used the app daily, completing eight CBT-based modules; measures of users’ impressions of the virtual agents, working alliance, user experience, and usability were assessed at a two-week follow-up. Separate 2 (Virtual Agent Animation Feature: Present vs. Absent) X 2 (Virtual Agent Conversational Feature: Present vs. Absent) X 2 (Depressive vs. Non-Depressive) ANOVAs were conducted to evaluate user impressions of the virtual agent, working alliance, and user experience. Additionally, a separate 2 (Virtual Agent Animation Feature: Present vs. Absent) X 2 (Virtual Agent Conversational Feature: Present vs. Absent) ANOVA was conducted to measure the system usability. The results revealed that animation and conversation did not significantly enhance user impressions or user experience of the virtual agent overall. However, users who were experiencing depressive symptoms found the virtual agent to be less knowledgeable and less novel compared to users who were not experiencing depressive symptoms. Interestingly, users experiencing depressive symptoms only found the virtual agent to be less novel when the agent was conversational. Furthermore, the findings demonstrated that animation, but not conversation, significantly impacted users’ working alliance with the agent. Unexpectedly, users with animated virtual agents reported less shared tasks and goals with the virtual agent than those with non-animated virtual agents, suggesting that animation hindered the user-agent working alliance in this context. Although animation led to negative outcomes for working alliance, animated virtual agents were able to increase usability ratings for AirHeart, specifically the users’ perceptions of mHealth app ease of use and satisfaction, system information arrangement, usefulness, and the usability of the app overall.
Recommended Citation
Hill, Victoria, "Investigating the Influence of Animation and Conversational Virtual Agent Features on the Usability and User Experience of the Smartphone Mental Health Application “Airheart”" (2025). All Theses. 4522.
https://open.clemson.edu/all_theses/4522
Author ORCID Identifier
0009-0004-7989-1105