Faster Qualitative Analysis of Autism Caregiving

Evidano is an AI-powered qualitative data analysis platform that accelerates transcription, auto-coding, cross-segment contrasts, and auditable code histories for interview and survey corpora.

This post demonstrates how to reproduce the July 17, 2026 PLOS One themes from 12 semi-structured interviews and an n=150 survey using an AI-enabled, auditable pipeline.

• The PLOS One dataset is small but richly coded, ideal for demonstrating thematic frequency, co-occurrence, and cross-segment contrasts.
• Follow a transparent 7-step runbook (ingest, transcribe with custom dictionary, auto-translate, import deductive nodes, audit codes, run contrasts, export reports) to reproduce results.
• Evidano features that matter here include custom dictionaries and PII redaction for transcription, AI-assisted auto-coding with versioned code histories, and co-occurrence networks for interpretation.
• The PLOS One study (published 17 July 2026) identified six interlocking themes that teams can use as deductive nodes then extend inductively.

The PLOS One paper (17 July 2026) documented 12 semi-structured interviews and a prior survey (n=150) and identified six interlocking themes shaping family quality of life.

Between neuroaffirmation and access: Parents’ experiences of autism caregiving in England documented 12 interviews and a prior survey and surfaced six themes that explain care ecology and access barriers.

Read the paper: PLOS One.

• Key payoff for analysts: the dataset is small but richly coded, ideal for demonstrating thematic, frequency, and cross-segment techniques.
• Context for teams: heavy administrative burden, diagnostic delays, cultural and gendered amplifiers, plus neuroaffirmation as a protective response.

The Methods section reports purposive recruitment from social media and community organisations, an online survey and 12 semi-structured interviews, verbatim transcription in NVivo, de-identification, and reflexive thematic analysis that produced six themes and subthemes.

• Reflexive thematic analysis (constructionist lens), iterative coding, analytic memos, no inter-coder reliability metric.
• Data limitations: English-only transcripts; small sample but diverse cultural contexts (including Muslim caregivers).
• Reproducibility cue: transcripts and repository access are restricted for confidentiality (University of Lincoln repository).

This 7-step runbook lists the inputs and outputs you need to reproduce the PLOS One themes using an AI-enabled, auditable pipeline.

Inputs: audio/video files, survey CSV (n=150), participant metadata (age, gender, ethnicity, diagnosis status). Desired outputs: theme counts, segment contrasts, co-occurrence network, quote bank for reports.

• 1) Ingest survey CSV and interview audio into Evidano; attach metadata fields (gender, ethnicity, diagnosis).
• 2) Transcribe with a custom dictionary (terms like 'EHCP', 'DLA', culturally specific words); enable PII redaction.
• 3) Auto-translate any non-English text using your glossary; review automated transcripts.
• 4) Import initial codebook (six themes from PLOS) as deductive nodes; run AI auto-coding to propose subcodes.
• 5) Audit and refine codes manually; write analytic memos inline and version the codebook.
• 6) Run thematic frequency tables, cross-segment contrasts (ethnicity × theme), and generate co-occurrence networks for interpretation.
• 7) Export visual reports and a searchable quote bank for stakeholders; produce an executive brief with evidence-linked recommendations.

The PLOS study (17 July 2026) shows how compact, well-collected qualitative data can illuminate access barriers and neuroaffirmation practices across diverse families.

• Try this on your corpus: upload transcripts and survey data to Evidano, run the 7-step runbook above, and generate an evidence pack for commissioners or service designers within days.
• Try Evidano for free to pilot a single theme (for example, 'Fighting the system(s)') and extract frequencies, co-occurrences, and a stakeholder-ready quote deck.