# Dis/Immersion in Mindfulness Meditation with a Wandering Voice Assistant

**Authors:** Bonhee Ku, Tatsuya Itagaki, Katie Seaborn  
**arXiv:** [2304.11286](https://arxiv.org/abs/2304.11286)  
**Date:** 2023-04-22  
**Summary page:** https://www.summarizepaper.com/en/paper/2304.11286v1/

## Key points

- Study explores use of voice-based virtual assistants (VAs) in mindfulness meditation
- Researchers developed "wandering voice" embodiment to address fixed location issue
- Multi-speaker VA embedded in yoga mat changes location based on meditation experience
- Qualitative user study conducted with twelve participants comparing fixed smart speaker to wandering VA embodied in yoga mat
- Thick descriptions from interviews revealed immersion and dis-immersion experiences during meditation sessions
- Findings provide design implications for "wandering voices" and propose new paradigm for VA embodiment
- Research paper published as extended abstract at 2023 CHI Conference on Human Factors in Computing Systems (CHI EA '23)

## Summary

This study explores the use of voice-based virtual assistants (VAs) in assisting people with mindfulness meditation. To address the issue of fixed location embodiment of VAs in smart speakers, smartphones and smart environments not providing intuitive support, the researchers developed a novel embodiment called the "wandering voice," which is co-located with the user and moves along with the meditation task. They created a multi-speaker VA embedded in a yoga mat that changes its location according to the meditation experience. To evaluate the effectiveness of this wandering VA embodiment, a qualitative user study was conducted involving twelve participants. The study compared the experience of using a typical fixed smart speaker to that of using the wandering VA embodied in the yoga mat. Thick descriptions from interviews with participants revealed both immersion and dis-immersion experiences during meditation sessions. The findings of this study provide design implications for "wandering voices" and propose a new paradigm for VA embodiment that can be extended to guidance tasks in various contexts. The research paper is published as an extended abstract at the 2023 CHI Conference on Human Factors in Computing Systems (CHI EA '23).