This post shows how to turn interviews and usability data from VR rehabilitation trials into actionable insights using AI-enabled qualitative analysis. The primary keyword “qualitative analysis of VR rehabilitation” guides practical steps for researchers and UX teams who need fast, defensible synthesis of interview transcripts, session logs, and clinical measures. The guidance below uses the July 30, 2026 PLoS One feasibility study of DizzyVR as a worked example and extracts concrete numbers, verbatim quotes, and reproducible analysis steps you can replicate with your own data.
Key Takeaways
According to the July 30, 2026 PLOS One study, a single-arm feasibility trial of DizzyVR collected eight-week session logs and semi-structured interviews and found high adherence and promising exploratory clinical changes (PLOS One).
The qualitative portion of the study used verbatim audio interviews transcribed and coded to identify usability, safety, and perceived clinical effect themes, giving a replicable example for AI-assisted qualitative workflows.
- 10 participants completed the DizzyVR intervention, out of 13 who started, between October 2024 and January 2025, producing a 76.9% completion rate according to the PLOS One report.
- Session adherence among completers was 91.25%, and 30% of participants reported brief simulator sickness events, per the PLOS One study published on July 30, 2026.
- Exploratory clinical measures in the PLOS One study showed Timed Up and Go (TUG) improvement from mean 9.53s to 7.34s (p = 0.002) and statistically significant gains in balance confidence (ABC, p = 0.007) and Functional Gait Assessment (FGA, p = 0.012).
- Participant quotations in the PLOS One interviews captured usability and safety perceptions, for example: "I felt safe. At no time did I think I was going to fall, " attributed to Participant 3 in the study.
What happened and how the study measured it
What happened: The PLOS One feasibility study evaluated DizzyVR in patients with vestibular disorders using eight weekly 50-minute immersive sessions and post-intervention semi-structured interviews (PLOS One).
How it was measured: The PLOS One study recorded objective kinematic logs (head and hand position once per second), session success rates per block, clinical measures (DHI, TUG, ABC, FGA), usability scores (SUS mean 68.5), satisfaction (USEQ mean 24.6), and simulator sickness (SSQ) across sessions.
Constraints and context: The PLOS One authors stressed this was a small, single-arm feasibility study (10 completers), so clinical outcomes are exploratory rather than confirmatory and require randomized trials for causal claims.
Findings snapshot
| Date / Period | Metric | Value (PLOS One) | Implication |
|---|---|---|---|
| Oct 2024–Jan 2025 | Participants started / completed | 13 started, 10 completed (76.9% completion) | Feasible recruitment and retention for a single-center pilot |
| Through intervention | Session adherence among completers | 91.25% average attendance | High in-session engagement supports longer trials |
| Published July 30, 2026 | Adverse events | 3 participants (30%) reported mild nausea/disorientation, 5 sessions affected (6.84% of sessions) | Mild simulator sickness that decreased over time; include ramp-up protocols |
| Pre vs Post (TUG) | Gait speed | Mean TUG 9.53s → 7.34s (p = 0.002) | Exploratory improvement in functional mobility worth testing in RCT |
| Post-intervention | Usability and satisfaction | SUS mean 68.5, USEQ mean 24.6 | Acceptable usability; some users expected more home autonomy |
Implications for researchers and clinical UX teams
Primary implication: The PLOS One feasibility data support using immersive, task-specific VR with objective kinematic logging and interview feedback to evaluate both usability and preliminary function in vestibular rehab.
Design decision: According to the PLOS One study, include objective performance metrics (head/hand tracking) and a semi-structured interview at endline to capture acceptability and safety narratives, because the study found that qualitative reports explained why SSQ scores fell over time.
Trial planning: The PLOS One authors recommend larger, randomized trials with longer follow-up to test durability; use the pilot’s adherence (91.25%) and effect sizes (TUG change 2.19s) for power calculations.
How Evidano helps translate DizzyVR interviews into robust qualitative evidence
Problem: Interview transcripts and session logs are time-consuming to analyze
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Solution mapping: The PLOS One study collected verbatim interviews and per-second kinematic logs; Evidano ingests audio or transcripts, aligns them with session logs, and generates thematic and frequency analyses to accelerate synthesis.
Problem: You need trustworthy transcripts and PII control
Evidano supports transcription with custom dictionaries and PII redaction and can import clinical audio directly, which matches the PLOS One method of recording interviews for verbatim coding.
Use case: Import recorded DizzyVR interviews, apply automated transcription, review named-entity redaction, and export cleaned transcripts for coding.
Problem: Linking qualitative themes to objective metrics
Evidano produces cross-segment and co-occurrence analyses that let you link themes (for example, "initial discomfort") to session-level metrics (for example, SSQ scores falling across sessions) as demonstrated in the PLOS One feasibility study.
Try the Evidano features page to see examples of visual outputs and exportable codebooks (Evidano features).
Problem: Clinicians need fast answers from mixed-methods pilots
Evidano surfaces quotable statements, counts, and representative extracts so you can present exact participant language such as the PLOS One quote: "I felt safe. At no time did I think I was going to fall" (Participant 3) alongside the numeric SSQ trend.
Operational step: Use Evidano to auto-generate a short findings brief for clinicians that pairs key quotes with the supporting metrics for rapid decisions.
Integrations and workflow
Evidano accepts transcript uploads, spreadsheets, and per-session CSV logs and can combine them into a single project; for audio-first pilots replicate the PLOS One approach by using Evidano’s transcription pipeline and then its thematic analysis tools.
If you need clinical-grade transcripts, consider Evidano speech-to-text for controlled transcription pipelines (Evidano speech-to-text).
FAQ: qualitative analysis of VR rehabilitation
How can AI speed up qualitative analysis of VR rehabilitation interviews?
Answer: AI automates transcription, initial code suggestions, and frequency counts so analysts can focus on interpretation.
Supporting detail: The PLOS One DizzyVR study used verbatim interview transcripts to extract usability and safety themes; an AI workflow replicates those steps and reduces manual coding time while preserving auditability.
What qualitative data should I collect in a VR feasibility study?
Answer: Collect semi-structured exit interviews, session-level success rates, and per-second tracking logs for mixed-methods synthesis.
Supporting detail: The PLOS One trial recorded head/hand kinematics once per second, session success rates per block, and semi-structured interviews, which together explained adherence and perceived safety.
Can AI identify safety signals such as simulator sickness?
Answer: Yes, AI can flag recurring symptom language and link it to SSQ scores or session timestamps for rapid triage.
Supporting detail: The PLOS One authors reported that SSQ scores decreased across sessions and that three participants reported nausea or disorientation; an AI pipeline can surface these patterns and the exact participant quotes that describe them.
How do I preserve rigor and transparency using AI for qualitative coding?
Answer: Use reproducible pipelines with versioned codebooks, reviewer reconciliation, and exportable audit trails.
Supporting detail: The PLOS One feasibility study used a codebook and manual coding; Evidano replicates that approach while logging every AI-assisted suggestion and each analyst decision for auditability.
Conclusion & Next Steps
The PLOS One DizzyVR feasibility study (published July 30, 2026) provides a clear mixed-methods template: capture objective kinematics, collect standardized clinical scales, and record semi-structured interviews to explain usability, safety, and perceived benefit.
AI-enabled qualitative analysis accelerates that template by automating transcription, proposing codes, quantifying theme frequencies, and linking quotes to session metrics for reproducible insights.
If you are designing a VR rehab pilot and want a tested workflow to analyze interviews and logs, start by aligning your interview guide with key metrics (SSQ, SUS, TUG) and then use an AI qualitative platform to scale coding and reporting.
Get started with a hands-on trial and see how this approach shortens synthesis time: Try Evidano for free.