On 15 July 2026 the Department for Science, Innovation and Technology published a 63-page report summarising parents' and children's views on social media, including attitudes to an under-16 ban and AI chatbots. The study draws on two nationally representative surveys (parents and children) conducted by Savanta plus a public consultation, raw inputs that are suitable for rapid qualitative analysis. Read the original report at Children's wellbeing online (final report). This post shows researchers and policy teams how to convert that corpus into thematic, segment-level, and visual insights using AI-enabled qualitative research, and how Evidano speeds the work without sacrificing privacy.
Key Takeaways
Evidano is an AI-powered qualitative data analysis platform that ingests PDFs, ODS tables, transcripts, and CSVs to produce thematic, cross-segment, and auditable insights for policy and research teams.
The DfSIT report published on 15 July 2026 combines two nationally representative surveys (parents and children) with open consultation annexes, and those annex ODS files are ideal for thematic coding and cross-segment comparison.
- Ingest the report PDF and Annex A–D ODS files to preserve respondent metadata and verbatim consultation responses for thematic and differential analysis.
- Combine representative polling (parents vs children) with thematic counts and exemplar quotes to produce decision-ready briefs that show both scale and texture.
- Use Evidano's automated theme extraction, co-occurrence visualizations, and clickable evidence to deliver auditable memos and stakeholder-ready visuals in days.
Fast take: what the report delivers
The report, published 15 July 2026, compiles quantitative polling and qualitative consultation about children's social media use, perceived harms and benefits, views on an under-16 ban, and attitudes toward AI chatbots and age assurance.
- Authors / publisher: Department for Science, Innovation and Technology; surveys run by Savanta.
- Primary inputs: 2 nationally representative surveys (parents; children and young people) plus public consultation annexes (multiple ODS tables).
- Why it matters: clear stakeholder split, parents and children often prioritise different harms, making cross-segment analysis essential for policy decisions.
Findings snapshot
| Date | Document | Sample / Inputs | Pages / Size | Core topics |
|---|---|---|---|---|
| 15 July 2026 | Children's wellbeing online (final report) | 2 nationally representative surveys (parents; children) + public consultation (Savanta) | 63 pages / PDF 765 KB | Device use patterns; benefits vs harms; under-16 ban; AI chatbots; age assurance; media literacy |
What the study did (plain English)
The study used representative polling for parents and for children and young people, supplemented by open consultation responses and targeted qualitative engagement, with annex tables available in ODS format.
- Scope: attitudes to platform-level bans (under-16), feature and time limits, AI chatbot risks and benefits, and age assurance views.
- Useful artifacts: raw consultation responses and annexed tables (Annex A–D), ideal for thematic coding, sentiment checks, and cross-segment comparisons.
Implications for researchers and policy teams
For qualitative researchers
Qualitative researchers should not treat the report as finished insight, because the public consultation contains verbatim responses that reveal nuance such as trade-offs parents accept versus children’s priorities.
Key action: compare code frequency and illustrative quotes across age bands and parent versus child samples to avoid policy blind spots.
For UX & product teams
UX and product teams should map the report's desired feature-level controls to user journeys, because the report highlights time limits and content filters that both parents and teens discuss.
Key action: run co-occurrence analysis (feature requests versus expressed harms) to prioritise low-friction interventions.
For policy analysts & decision makers
Policy analysts and decision makers should weigh population-level polling against qualitative submissions, because representative metrics support headline decisions while consultation responses reveal likely unintended consequences such as circumvention and migration to unregulated platforms.
Key action: weigh representative metrics against thematic evidence from consultations when modelling outcomes for an under-16 ban.
Do more, faster with Evidano
Ingest any format, preserve provenance
Evidano ingests PDFs, ODS tables, interview transcripts, and open consultation CSVs into one workspace while preserving provenance.
Why it matters: the report includes annex ODS files; ingesting them preserves respondent metadata such as age band and respondent type for cross-segment analysis.
Automated thematic + cross-segment analysis
Evidano generates themes, subthemes, and frequency counts across segments, letting you quantify how often mental health or age assurance appears by cohort.
Use case: measure support for an under-16 ban by combining representative polling with thematic sentiment from open responses.
Clickable evidence & visualizations
Evidano exports co-occurrence networks, hierarchical code trees, and quote packs while preserving links back to original verbatim responses so stakeholders can drill from a theme to source evidence.
Benefit: faster, verifiable memos for policy teams and publishable visual assets for stakeholder meetings.
Secure, research-first handling
Evidano encrypts data and does not use customer data to train external models, and the platform includes redaction tools for PII in sensitive child and parent data.
Ethics note: these tools are for research and policy analysis only, not for clinical diagnosis.
7‑step workflow: from report to decision-ready insight
This 7-step workflow takes about a week to reproduce and extend the DfSIT analysis using AI-assisted tools.
- 1) Collect assets: download the PDF and Annex A–D ODS files from the report (see Children's wellbeing online (final report)).
- 2) Ingest: upload PDFs and ODS to Evidano; map respondent metadata such as age band and parent or child.
- 3) Auto-code: run initial theme extraction to surface the top 20 themes across the corpus.
- 4) Refine codebook: import or edit codes and apply supervised coding to ensure consistency.
- 5) Cross-segment analysis: run frequency and differential sentiment reports comparing parents versus children and age bands.
- 6) Visualize & extract quotes: generate a co-occurrence network and prepare a one-page brief with exemplar quotes.
- 7) Iterate with stakeholders: use Evidano's document chat to answer follow-ups and generate policy option memos.
FAQ: qualitative analysis of children's social media
How do I compare representative polling with open consultation responses?
Use the polling for population-level context and the consultation responses for depth, because combining both shows alignment and divergence between scale and texture.
Practical step: tag each source in your workspace and run side-by-side theme frequency and sentiment reports to highlight where representative metrics and verbatim responses agree or differ.
Can sensitive child data be handled safely?
Yes, sensitive child data can be handled safely by redacting PII, limiting user access, and exporting redacted quote packs for publication.
Practical step: follow institutional ethics approvals and use built-in redaction and access controls, remembering that these analyses are research-focused, not diagnostic.
What's the quickest win for a policy brief?
The quickest win is a short brief that combines representative poll metrics with the top three qualitative themes and illustrative quotes, because that format provides both scale and texture for decisions.
Practical step: produce a one-page decision brief with headline poll figures and three exemplar quotes drawn from the consultation annexes.
Wrapping up: start turning the report into action
The DfSIT 15 July 2026 report is a rich mixture of representative metrics and qualitative submissions, ideal for AI-enabled thematic and cross-segment research.
- Try a reproducible pilot: ingest the report's PDF and annex ODS files, run theme extraction, and produce a one-page decision brief in days, not weeks.
- Get started: Try Evidano for free or visit Evidano to discuss a pilot for your policy or research project.