The primary keyword for this post is qualitative analysis help-seeking criminal justice. Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. According to the PLOS ONE article, help-seeking for mental health and substance use among people in contact with the criminal justice system is shaped by individual desire, ability, and the social context in which people are embedded. This post translates the PLOS ONE (Connell et al., 2026) findings into concrete methods and AI-enabled workflows that qualitative researchers and policy teams can adopt to surface relational influences from interview transcripts and network-assisted interviews.
Key Takeaways
According to the PLOS ONE article (Connell et al., 2026), relational and contextual influences across social networks are central to whether people under community criminal justice supervision seek help for mental health and substance use, and these influences cannot be reduced to individual factors alone. PLOS ONE
- The PLOS ONE study interviewed 50 people between 1 November 2023 and 31 July 2024 and published the qualitative findings on 30 July 2026.
- According to PLOS ONE (Connell et al., 2026), participants came from two contrasting local authorities: 23 in LA1 and 27 in LA2, demonstrating context-specific differences in service access.
- Background prevalence cited in the article notes over 40% meet diagnostic criteria for a substance use disorder, more than 10% for a major depressive episode or psychotic disorder, and 75% for any personality disorder in related probation populations (as reviewed by Connell et al., 2026).
- A core participant quote in the PLOS ONE article summarized relational constraints: "you just kind of can’t get away from the social, every time you try and do something good basically they will get you on it and you will be wrecked and miss what you are going to do" (Participant P13, Connell et al., 2026).
What Happened: Methods and main finding
Answer: The PLOS ONE team conducted an in-depth qualitative framework analysis of 50 interviews to identify how desire, ability, and context interact to shape help-seeking among people supervised by community criminal justice services.
According to the PLOS ONE article (Connell et al., 2026), researchers used semi-structured, audio-recorded interviews aided by Network Canvas software and transcribed interviews verbatim; interviews took place from 1 November 2023 to 31 July 2024 and were analyzed using framework analysis in NVivo.
The PLOS ONE study confirmed that help-seeking is a relational process: individual attitudes and capacities interact with network beliefs, resources, and local service availability to determine whether help is sought and sustained.
Researchers and commissioners should treat help-seeking as a multi-level target for intervention, because the PLOS ONE findings show that changing individual attitudes without addressing network norms and service trust is unlikely to succeed.
Findings Snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| 1 Nov 2023–31 Jul 2024 | Interviews conducted | 50 total (23 in LA1, 27 in LA2) | Sample size supports rich qualitative framework and cross-context comparison |
| 30 Jul 2026 | Publication | PLOS ONE article (Connell et al., 2026) | Findings available for policy and research uptake |
| Background (cited in article) | Substance use disorder prevalence | Over 40% (in related probation samples) | High unmet need in community-supervised populations |
| Background (cited in article) | Personality disorder prevalence | Approximately 75% (in related samples) | Complex comorbidity that affects engagement with services |
Implications for qualitative researchers and policy teams
Answer: The PLOS ONE framework implies that qualitative researchers should explicitly capture social-network and contextual data alongside individual narratives to understand help-seeking.
According to Connell et al. (PLOS ONE, 2026), adding a network map and targeted prompts about network attitudes during interviews reveals shared norms, epistemic mistrust, and resource flows that standard one-to-one interviews miss.
Practical decisions: in future studies plan to collect network composition (ties, duration, perceived support), code relational themes (endorsement, sabotage, normalization), and compare across contexts as the PLOS ONE study did between LA1 and LA2.
Policy teams should prioritise multi-level interventions: the PLOS ONE authors recommend combining individual outreach with community-level trust-building and service redesign to reduce exclusionary access rules.
How Evidano Helps
Problem: Interviews contain relational signals that are hard to surface at scale
Solution: Evidano automates thematic and cross-segment analysis to extract network-related themes such as 'network endorsement' or 'epistemic mistrust' from transcripts and linked metadata.
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. For teams that collect network-mapping data like Network Canvas, Evidano can ingest transcripts and structured network spreadsheets to produce combined thematic and frequency outputs.
Problem: Synthesis across contexts is time consuming
Solution: Evidano generates side-by-side code frequency and cross-segment comparisons so you can compare LA1 vs LA2 patterns (for example, confidentiality concerns in small towns vs service overload in urban areas).
See Evidano Features for thematic coding, co-occurrence networks, and cross-segment analyses that match the multi-level framework recommended by Connell et al. (PLOS ONE, 2026).
Problem: Teams need secure, reproducible workflows for sensitive qualitative data
Solution: Evidano supports encrypted storage, transcript pseudonymisation, PII redaction, and research reproducibility via exportable codebooks and visualisations, enabling ethical reuse aligned with the PLOS ONE data-access note.
FAQ: qualitative analysis help-seeking criminal justice
How can AI help qualitative analysis of help-seeking in criminal justice contexts?
Answer: AI can accelerate coding, surface network-related co-occurrence patterns, and produce cross-segment comparisons so researchers can test the PLOS ONE framework at scale.
AI-assisted platforms can auto-suggest codes for relational themes (for example, 'mistrust of services' or 'peer endorsement') and compute code frequencies by subgroup, which speeds iterative framework refinement as demonstrated in the PLOS ONE study.
Which data should I collect to apply the PLOS ONE framework?
Answer: Collect verbatim interview transcripts, basic network maps (ties, relationship type, duration), and contextual site metadata such as local authority and service availability.
The PLOS ONE study combined interview transcripts with participant-aided sociograms, which enabled identification of network norms and resource flows; replicating this requires both qualitative text and simple network spreadsheets.
Is this approach ethical for sensitive populations?
Answer: Yes, when researchers apply strict consent, pseudonymisation, and controlled access procedures as described in the PLOS ONE article and its UK Data Service deposit.
Connell et al. (PLOS ONE, 2026) obtained informed consent for pseudonymised data reuse and stored data with access controls; researchers should mirror those safeguards and consider a data access committee for sensitive datasets.
What practical outputs should funders expect from this methodology?
Answer: Funders should expect a coded framework of desire, ability, and context, cross-site comparisons, and actionable recommendations for multi-level interventions.
The PLOS ONE framework is designed to inform individual, network, and community-level interventions and can be operationalised into measurable targets such as increases in network-level trust or reduced service exclusion criteria.
Conclusion & Next Steps
The PLOS ONE (Connell et al., 2026) study shows that help-seeking among people in contact with the criminal justice system is shaped by desire, ability, and the relational context embedded in social networks.
Researchers should collect network-aware qualitative data and use AI-enabled tools to synthesise relational themes across contexts, following the mixed-methods approach used in the PLOS ONE article.
Evidano helps teams operationalise this: ingest transcripts and network data, run thematic and cross-segment analyses, and produce reproducible codebooks and visualisations to inform multi-level interventions.
If you want to pilot an AI-enabled workflow for network-informed qualitative research, Try Evidano for free.