Problem: The SOUND-BITES pilot embeds hearing checks into Meals on Wheels visits and will collect optional participant and volunteer interviews to evaluate acceptability and feasibility (target: 60 volunteers, 600 clients; recruitment started 01/03/2026). What you’ll learn: a compact, reproducible workflow for qualitative analysis of hearing program interviews and how to run it with Evidano. The payoff: turn interview audio recorded in homes (with variable noise and formats) into coded themes, cross-site comparisons, and stakeholder-ready visuals while keeping data private. Primary keyword: qualitative analysis of hearing program interviews. Source: full protocol (published 14 July 2026) at PLoS One.
Findings snapshot
| Date / Milestone | Metric | Value | Source | Implication |
|---|---|---|---|---|
| Published | Article date | 14 July 2026 | PLoS One | Protocol available for method mapping |
| Ethics | Approval date | 11 Dec 2025 (Macquarie HREC ID: 16818) | Protocol | Confirms interview procedures and consent |
| Recruitment | Start | 01/03/2026 (students) | Protocol | Student placements in progress; client recruitment pending |
| Sample target | Volunteers & clients | 60 volunteers; 600 clients/household members | Protocol | Potentially large mixed-method corpus |
| Qual data | Interview timing | 4 weeks post-program (optional) | Protocol | Interviews for acceptability/feasibility; saturation expected 9–17 |
| Tools | mHealth devices | Sound Scouts app; Arclight otoscope | Protocol | Generates clinical outputs to triangulate with interviews |
What happened and what you’ll analyze
The SOUND-BITES pilot trains Meals on Wheels volunteers and audiology students to deliver in-home otoscopy and an 8–10 minute Sound Scouts hearing screen, then offers optional recorded interviews with clients and volunteers to probe acceptability and feasibility. Interviews are scheduled about 4 weeks after the visit and include structured and open-ended prompts tied to the Theoretical Domains Framework (TDF).
- Audio: phone or Zoom / telephone; intelligent verbatim transcription in Word; manual checking per protocol.
- Qual coding plan in protocol: deductive content analysis mapped to TDF, dual coders with consensus and third-party adjudication.
- Analysis outputs: themes for acceptability, barriers/facilitators, help-seeking decisions; triangulation with survey and clinical outputs.
Why AI matters for qualitative analysis of hearing program interviews
Problem → Opportunity
Interviews recorded in home environments have variable audio quality and inconsistent metadata, which make manual workflows slow and error-prone at scale. Manual verbatim workflows (transcribe, clean, code in NVivo) are slow and error-prone when n scales toward dozens or hundreds of interviews. AI-enabled tools reduce manual steps: automated, reviewable transcripts; accelerated code suggestions; frequency and cross-segment summaries that surface patterns across sites, volunteer cohorts, or hearing-number outcomes.
Do more, faster with Evidano
Evidano accelerates transcription, TDF-aligned coding, cross-segment analysis, and secure reporting for qualitative hearing interviews. Transcription: AI transcription with custom dictionary (technical terms like "Arclight", "Sound Scouts", hearing-number ranges) and PII redaction to meet consent protocols.
Coding & themes: import codebook (TDF) and run AI-assisted coding to pre-label passages for dual-coder review, reducing initial coding time by orders of magnitude.
Cross-segment analysis: automatic thematic frequency and co-occurrence analysis by site, volunteer experience, or referral outcome to spot which sites show more help-seeking.
Triangulation: link interview themes to Sound Scouts outputs and survey HHIE‑S scores to quantify which qualitative barriers predict non-help-seeking.
Visuals & export: generate word clouds, co-occurrence networks, hierarchical theme to subcode trees and exportable quotes for stakeholder briefs.
Privacy & governance: Evidano uses encrypted storage and proprietary LLMs tuned for qualitative research, data is never used to train third-party models.
Where Evidano saves time on the SOUND-BITES workflow
Evidano reduces time across the SOUND-BITES workflow by automating transcripts, pre-labeling codes, and generating reports. From import to insights in days not weeks: automated transcripts, AI-assisted code suggestions, cross-segment reports and slide-ready summaries.
Reduce rework: reviewer-friendly transcript editors and clickable source audio let coders verify quotes without switching apps.
Make audit trails simple: timestamped coding decisions and exportable codebooks for ethics and publication.
Workflow: reproduce the protocol’s qualitative analysis in 7 steps
Follow this checklist to run qualitative analysis of hearing program interviews at scale.
- 1) Import audio files and metadata (site, volunteer ID, client age, Sound Scouts hearing number) into Evidano.
- 2) Run automated transcription with a custom dictionary (Arclight, Sound Scouts, hearing-number ranges) and enable PII redaction. Review and correct transcripts.
- 3) Import the TDF codebook used in the protocol and run AI-assisted pre-coding to surface candidate segments per domain.
- 4) Dual-code a subset (n≈10–20%) to calibrate, lock codes, apply AI labeling across corpus, reviewers verify and finalize.
- 5) Run cross-segment analyses (by site, by volunteer experience, by referral outcome) and co-occurrence networks to detect linked barriers/facilitators.
- 6) Triangulate with quantitative survey and audiometric outputs (HHIE‑S, hearing-number, referrals) to prioritise action items.
- 7) Export concise executive brief: top 5 themes, representative quotes with timestamps, visuals (co-occurrence map, theme frequency table), and reproducible analysis files for publication.
FAQ: qualitative analysis of hearing program interviews
How many interviews do I need for saturation?
The SOUND-BITES protocol cites saturation estimates of 9–17 interviews for thematic saturation. For cross-segment comparisons (by site or volunteer subgroup), increase interviews per subgroup (≥12) to detect meaningful differences.
How do I compare segments reliably?
Use Evidano to tag each transcript with structured metadata (site, volunteer role, hearing-number category, referral action). Run frequency, chi-squared summaries and co-occurrence analyses to quantify differences and export 95% CI tables for binary outcomes.
Is AI transcription valid in noisy, in-home audio?
Noisy recordings require higher-review effort: enable noise-aware models, use the platform’s confidence scores to prioritize manual checks, and keep the protocol’s manual verification step to ensure fidelity.
Data ethics & research note
This guidance supports research workflows only. Follow the SOUND-BITES consent and HREC conditions (Macquarie HREC ID: 16818). Evidano provides PII redaction and secure storage, do not use outputs for clinical diagnosis.
Wrapping up: next steps
The next steps are to map audio, survey, and clinical outputs into a single, auditable corpus and apply the 7-step workflow above to accelerate synthesis and publication. Try a pilot: ingest 10 interviews, enable custom-dictionary transcription, import the TDF codebook, and run an AI-assisted code pass to see savings firsthand.
Ready to try it with your SOUND-BITES data? Try Evidano for free to see how Evidano handles transcription, TDF-aligned coding, cross-segment analysis and secure reporting for your study.