MétaCan
Menu
Back to cohort
Record W4413849445 · doi:10.2196/71004

Avatar Customization and Embodiment in Virtual Reality Self-Compassion Therapy for Depressive Symptoms: Three-Part Mixed Methods Study

2025· article· en· W4413849445 on OpenAlexvenueno aff
T.C. Elliott, Jarrod Knibbe, Julie D. Henry, Nilufar Baghaei

Bibliographic record

VenueJMIR Formative Research · 2025
Typearticle
Languageen
FieldComputer Science
TopicVirtual Reality Applications and Impacts
Canadian institutionsnot available
Fundersnot available
KeywordsAvatarPreprintSelf-compassionPsychologyPersonalizationMixed realityVirtual realityPsychotherapistComputer scienceHuman–computer interactionWorld Wide WebMindfulness

Abstract

fetched live from OpenAlex

Background: As virtual reality technologies become more accessible, understanding how design features influence user experience (UX) and psychological benefit is critical, particularly for emotionally sensitive interventions. Thus, while prior studies support the use of self-compassion paradigms in immersive virtual reality (VR) environments, the effects of avatar stylization, customization, and mirrored self-representation on therapeutic outcomes are not well understood. For instance, while it is often assumed that increasingly realistic avatars are preferable to less realistic ones, this basic premise remains largely untested. Objective: This study aimed to evaluate whether avatar appearance, customization features, and virtual mirrors affect UX and therapeutic outcomes in VR self-compassion therapy. Specifically, we examined whether stylized avatars, avatar customization, and virtual mirror feedback influenced user-rated self-compassion and depression symptoms. Methods: Across three between-subjects studies (N=107 neurotypical adults), participants engaged in an immersive individualized VR therapy protocol based on a 2-phase compassion task. The conditions were (1) stylized avatars (n=20), (2) stylized customizable avatars (n=49), and (3) stylized customizable avatars with a virtual mirror (n=38). Participants completed the User Experience Questionnaire, the Self-Compassion Scale, and the 8-item Patient Health Questionnaire (PHQ-8). In study 3, presence was also assessed using the Slater-Usoh-Steed scale. Qualitative feedback was analyzed thematically. Between- and within-study comparisons used t tests and Mann-Whitney U tests. Results: Avatar customization (study 2) led to a significant increase in self-compassion (Self-Compassion Scale: baseline mean 3.05, SD 0.98; follow-up mean 3.55, SD 1.16; t89=2.22; P=.03; d=-0.47), though PHQ-8 scores remained unchanged. The virtual mirror condition (study 3) significantly improved depression scores (PHQ-8: U=477.5; z=2.53; P=.01; r=0.30) and UX across four User Experience Questionnaire categories, including attractiveness and dependability. However, self-compassion did not significantly change in study 3 (mean 3.88, SD 1.33 → mean 4.09, SD 1.05; t63=0.71; P=.47; d=0.18). Presence scores in study 3 (mean 4.56, SD 1.58) were also comparable to real-world benchmarks. Qualitative feedback highlighted strong engagement with avatars and mirrors, and participants reported emotional safety and personalization benefits. Conclusions: Stylized avatars, when paired with customization and mirrored embodiment, can support UX and therapeutic benefit in VR self-compassion therapy. These findings challenge the assumption that hyperrealistic avatars are superior and highlight the importance of emotionally congruent design choices. The combination of stylization, individualization, and visual feedback may offer a low-barrier, user-aligned strategy for future therapeutic VR applications.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.005
metaresearch head score (Gemma)0.005
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Qualitative · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.005
Threshold uncertainty score0.025

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0050.005
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0010.000
Science and technology studies0.0010.001
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0030.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.068
GPT teacher head0.469
Teacher spread0.401 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designQualitative
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations1
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueJMIR Formative ResearchSame topicVirtual Reality Applications and ImpactsFrench-language works237,207