Alexa depression and anxiety self-tests

Abstract

Mental health resources available via websites and mobile apps provide support such as advice, journaling, and elements from cognitive behavioral therapy. The proliferation of spoken conversational agents, such as Alexa, Siri, and Google Home, has led to an increasing interest in developing mental health apps for these devices. We present the pilot study outcomes of an Alexa Skill that allows users to conduct depression and anxiety self-tests. Ten participants were given access to the Alexa Skill for two-weeks, followed by an online evaluation of the Skill's usability and trust. Our preliminary evaluation suggests that participants trusted the Skill and scored the usability and user experience as average. Usage of the Skill was low, with most participants using the Skill only once. In view of work-in-progress, we also present a discussion of implementation and study design challenges to guide the current literature on designing spoken conversational agents for mental health applications. CCS CONCEPTS • Human-centered computing → Sound-based input / output.

Amazon Echo Show as a Multimodal Human-to-Human Care Support Tool within Self-Isolating Older UK Households

“…Building on Isaac's discussion of his own isolation too, there are opportunities for VAs to do more than simply leveraging the use of Alexa 'skills' to handle occasions where people speak to the device when they are experiencing high anxiety [62]. This reflects Roth's description of how engagement technology can move from the individual to the wider, social environment [66].…”

Voice-based conversational agents for sensing and support: Examples from academia and industry

“…No case presented pure proactive support. Four cases presented a blended service: Lifepod, WellBe, and Pria provide a mix of active sensing and proactive support, Aiva offers a combination of reactive and proactive support (e.g., information lookup, reminders, alerts), and Quiroz and colleagues (Quiroz et al, 2020) presented a prototype offering active sensing and reactive support.…”

“…Furthermore, while companies deliver solutions that are not illness-specific but that can be applied to both mental health and substance use disorders, academia tends to focus on selected conditions. In particular, academic research seems to focus either on evaluating existing commercial VCA for the support of specific health conditions (Boyd & Wilson, 2018;Miner et al, 2016;Nobles et al, 2020;Reis et al, 2018;Yang et al, 2021) or on developing Alexa skills for the same purpose (Cheng et al, 2018;Maharjan et al, 2019;Motalebi & Abdullah, 2018;Quiroz et al, 2020;Wang, Yang, Shao, Abdullah, & Sundar, 2020). In only a few of the reviewed cases, independent VCA prototypes were developed (Adaimi et al, 2020;Quiroz et al, 2020;Swamy et al, 2019).…”

“…In particular, academic research seems to focus either on evaluating existing commercial VCA for the support of specific health conditions (Boyd & Wilson, 2018;Miner et al, 2016;Nobles et al, 2020;Reis et al, 2018;Yang et al, 2021) or on developing Alexa skills for the same purpose (Cheng et al, 2018;Maharjan et al, 2019;Motalebi & Abdullah, 2018;Quiroz et al, 2020;Wang, Yang, Shao, Abdullah, & Sundar, 2020). In only a few of the reviewed cases, independent VCA prototypes were developed (Adaimi et al, 2020;Quiroz et al, 2020;Swamy et al, 2019). These were mainly intended as proof-of-concept solutions, which were not made available on the market.…”

“…Nine of the presented cases describe a voice-based intervention delivered through commercial VCAs, using either Google Actions or Alexa Skills. Seven of those nine used Skills, whereas five were coming from academia (Cheng et al, 2018;Maharjan et al, 2019;Motalebi & Abdullah, 2018;Quiroz et al, 2020;Wang, Yang, Shao, Abdullah, & Sundar, 2020), and two were developed in industry (Headspace, 2021b;Jar, 2021). Only in a few cases, Google Actions were used (Calm, 2020;Headspace, 2021a;Infermedica, 2021;Jar, 2021).…”