AI Qualitative Analysis: Lessons from ELLY

Researchers evaluating community health interventions need to understand more than whether an outcome changed. Researchers also need to explain how participants experienced an intervention, which mechanisms supported change, and where local barriers limited engagement. AI qualitative analysis can help public health, UX, and policy research teams organize those complex interview findings while preserving the evidence behind each theme. Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and research documents. The November 5, 2026 PLOS One study of Enjoy Life LocallY, or ELLY, provides a useful example: researchers interviewed 20 participants from two disadvantaged Scottish communities after a 12-week intervention delivered to 75 residents. The study used framework analysis to identify pathways involving social connection, behavior change, diet, wellbeing, and new skills. This article explains what the ELLY findings show, how researchers can use AI-assisted thematic analysis responsibly, and how Evidano can accelerate synthesis without replacing methodological judgment.

Key Takeaways

The PLOS One study found that ELLY was acceptable and feasible in two disadvantaged Scottish communities, with participant interviews suggesting that wellbeing and social connection were more closely associated with behavior change than weight-focused goals.

In September 2023 to November 2023, the ELLY intervention served 75 residents through free soup, community activities, goal setting, self-monitoring, and a loyalty card offering a £25 shopping card after at least 9 activities over 12 weeks.
In November 2026, the researchers reported qualitative interviews with 20 participants, 10 from each community, using semi-structured interviews and framework analysis.
The November 2026 study identified interconnected pathways involving stronger social connectedness, healthier dietary habits, positive behavior changes, improved participant-reported wellbeing, and new skills and knowledge.
The ELLY researchers stated that qualitative evidence alone was insufficient to establish definitive causation, making the findings useful for understanding perceived mechanisms rather than proving intervention effectiveness.
AI qualitative analysis can help research teams compare themes across communities, trace participant quotations to codes, and identify barriers that may be missed in manual synthesis.

How AI Qualitative Analysis Clarifies the ELLY Findings

The ELLY study shows that qualitative health research analysis is most valuable when it connects participant experience to intervention mechanisms and outcomes. According to the November 2026 PLOS One study, researchers conducted approximately 30-minute, face-to-face interviews at 12 weeks and used field notes, team meetings, and independent coding checks to strengthen interpretation.

The study design combined community co-design with a 12-week feasibility intervention delivered from September 2023 through November 2023. According to the ELLY researchers, the intervention included free soup twice weekly, access to local activities and groups, optional goal setting, written and online information, and a loyalty card that rewarded engagement.

The researchers used interviews to explore how participants experienced the intervention rather than to calculate an independent treatment effect. The PLOS One authors wrote that the qualitative data offered “rich and valuable insight into participants’ experiences and perceived pathways to observed outcomes.”

The resulting logic model separated mechanisms from outcomes. According to the November 2026 article, snowball recruitment and community connections were mechanisms, while social connectedness and positive behavior change were outcomes linked through multiple, overlapping pathways.

AI-assisted analysis can make this structure easier to inspect by grouping recurring concepts, comparing themes between communities, and surfacing contradictory cases. Human researchers must still review the coding framework, context, positionality, consent constraints, and causal limits.

ELLY Findings Snapshot for Health Researchers

Date or period	Metric	Value	Implication for qualitative analysis
September 2023 to November 2023	Intervention duration	12 weeks	Researchers could examine immediate and intermediate participant experiences, but not long-term maintenance.
September 2023 to November 2023	Residents receiving the intervention	75	The feasibility study provided a defined population for interpreting interview recruitment and reach.
November 2026	Interview participants	20, with 10 per community	The interview sample supported comparison between two local contexts.
November 2026	Interviewees represented among all intervention participants	27%	The qualitative sample was substantial for a feasibility study but did not include the 9 participants who dropped out.
November 2026	Participants reporting comorbidities	12 of 20, or 60%	Health and wellbeing narratives reflected varied personal circumstances.
November 2026	Participants living in disadvantaged areas	17 of 20, or 85%	The findings were closely tied to the study's target population and local context.
November 2026	Female interview participants	14 of 20, or 70%	Researchers should consider sample composition when assessing transferability.

Implications of AI Qualitative Analysis for Public Health Researchers

How can researchers distinguish intervention mechanisms from outcomes?

Researchers can distinguish mechanisms from outcomes by coding what an intervention changed in participants' environments or motivations separately from what participants reported afterward.

The November 2026 ELLY study treated snowball recruitment, community connections, and access to activities as pathways or mechanisms, while treating social connectedness, positive behavior change, and improved wellbeing as outcomes. An AI qualitative analysis workflow can preserve those distinctions through hierarchical codes and relationship analysis, but researchers must validate the conceptual model against the original transcripts.

Why should health researchers analyze negative cases as well as positive themes?

Health researchers should analyze negative cases because the ELLY interviews showed that the intervention did not work in the same way for every participant.

According to the November 2026 PLOS One study, some participants found community groups unwelcoming, difficult to approach, irregular, or poorly communicated. AI-assisted cross-segment analysis can flag these less frequent but consequential experiences alongside dominant themes.

What does the ELLY study suggest about weight-focused interventions?

The ELLY study suggests that wellbeing and social connection may motivate behavior change more directly than healthy weight goals alone.

The November 2026 researchers reported that healthy weight was not perceived as the primary motivator for many participants and described weight-related improvement as a possible “happy by-product” of better wellbeing, social connection, and healthier routines. The authors also cautioned that the qualitative findings did not establish definitive causation.

How can AI help researchers compare disadvantaged communities?

AI can help researchers compare communities by applying a consistent coding structure across transcripts while retaining community-specific language and context.

The November 2026 ELLY article reported broadly comparable findings across two communities but also identified differences in community cohesion, activity availability, communication, and local partner engagement. Cross-segment analysis can make these similarities and differences visible earlier in the research process.

How Evidano Supports AI Qualitative Analysis of Health Interviews

Problem: Interview synthesis is slow and difficult to audit

Evidano is an AI-powered qualitative data analysis platform that helps researchers organize interview transcripts, research papers, reports, and survey responses into evidence-linked findings.

Evidano supports thematic, content, frequency, and cross-segment analyses, which can help a public health team examine themes such as social connection, diet, confidence, and barriers across communities. Researchers can review the source material and use Evidano's features to move from raw qualitative data to a structured analytic view.

Problem: Researchers need to compare themes across locations or participant groups

Evidano supports cross-segment analysis so researchers can compare how themes appear across communities, demographic groups, intervention components, or other study segments.

For an ELLY-style study, a research team could compare Community C1 with Community C2, examine positive and negative engagement experiences, and identify whether social connection or activity access appears differently across local contexts.

Problem: Researchers need to preserve sensitive health data controls

Evidano encrypts research data and does not use customer data to train third-party models, according to Evidano's data-handling policy.

Health researchers should still follow their institutional ethics approvals, consent requirements, data minimization practices, and secure access procedures. Details about platform safeguards are available on Evidano's data security page.

Problem: Researchers need to interrogate a complex logic model

Evidano's AI chat enables researchers to ask questions across uploaded documents and analyses, while visualizations can show code hierarchies and co-occurrence patterns.

A researcher could use this workflow to ask which quotations support a proposed pathway, where participants described barriers, and whether a mechanism appeared in both communities. Human review remains necessary before researchers treat an AI-generated pattern as an analytic conclusion.

FAQ: AI Qualitative Analysis in Community Health Research

What is AI qualitative analysis?

AI qualitative analysis is the use of artificial intelligence to organize, code, compare, summarize, and query non-numerical research data such as interviews, documents, and open-ended survey responses.

AI qualitative analysis can accelerate discovery of recurring themes and differences, but the November 2026 ELLY study demonstrates why researchers must retain human oversight of context, interpretation, reflexivity, and causal claims.

How many people were interviewed in the ELLY qualitative study?

The ELLY researchers interviewed 20 participants, with 10 participants from each of two disadvantaged Scottish communities, according to the November 2026 PLOS One article.

The 20 interviewees represented 27% of the 75 residents who received the intervention, and the study did not include the 9 participants who dropped out.

What qualitative analysis method did the ELLY researchers use?

The ELLY researchers used framework analysis to analyze semi-structured participant interviews, according to the November 2026 PLOS One study.

Two researchers developed an initial coding framework from five interviews, the wider team refined the framework, 10% of the data was independently cross-checked, and disagreements were resolved through discussion.

What were the main themes in the ELLY interviews?

The main ELLY interview themes were community connection, positive behavior change, healthier diet, mental and physical health improvements, and new skills and knowledge.

The November 2026 researchers also identified snowball recruitment, local asset awareness, inclusion barriers, irregular activities, and communication problems as important parts of the participant experience.

Can AI qualitative analysis prove that a health intervention caused an outcome?

AI qualitative analysis cannot by itself prove that a health intervention caused an outcome.

The ELLY researchers explicitly stated in November 2026 that qualitative data alone was insufficient to establish definitive causation. AI can help map perceived pathways and counterexamples, while causal claims require an appropriate study design and complementary evidence.

Conclusion and Next Steps for AI Qualitative Analysis

The November 2026 ELLY study shows how qualitative interviews can explain why a community health intervention was acceptable, how participants experienced change, and where local conditions limited inclusion. The study's 20 interviews from two communities identified interconnected pathways involving social connection, wellbeing, behavior, diet, and confidence, while also documenting the limits of short-term qualitative evidence. AI qualitative analysis can help researchers manage this complexity through thematic coding, cross-segment comparison, source-grounded questions, and visual analysis. Evidano gives health and policy research teams a structured way to turn interview data into reviewable findings while keeping human interpretation at the center. Try Evidano for free to analyze qualitative research data more efficiently.

Topics

AI qualitative analysis
AI qualitative data analysis
qualitative health research analysis
thematic analysis of interviews
community health research

Keep reading

Browse all articles