Researchers and applied qualitative teams face a common problem: help-seeking for mental health and substance use among people supervised in the community is shaped by relationships and context, not just individual choice. The primary keyword for this post, qualitative analysis help-seeking criminal justice, signals the focus: methods and takeaways for researchers who must surface relational drivers from interviews and network data. According to Connell et al., in a PLOS ONE article published 30 July 2026, social and network influences dominate help-seeking decisions in this population. Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. Ethics note: this post is research-focused and non-diagnostic.
Key Takeaways
According to the PLOS ONE article by Connell et al. (published 30 July 2026) PLOS ONE, help-seeking for mental health and substance use among people in contact with the criminal justice system is driven by interacting individual, relational, and contextual factors rather than being solely an individual choice. The PLOS ONE study found relational influences operate across social networks and local culture and that interventions should therefore span individual, network, and community levels.
- The PLOS ONE study interviewed 50 people between 1 November 2023 and 31 July 2024 and reported recruitment from two local authority areas with 23 participants in LA1 and 27 in LA2 (Connell et al., PLOS ONE, 30 July 2026).
- The study reports that 49 of 50 participants consented to verbatim quotation for research reuse in the UK/EU (Connell et al., PLOS ONE, 30 July 2026).
- Connell et al. (PLOS ONE, 30 July 2026) state plainly, “help-seeking is not an individual behaviour, but strongly affected by relational influences, ” highlighting network-level mistrust, stigma, and resource flows as core barriers.
- Interviews ran up to two hours, used participant-aided sociograms and Network Canvas for social network elicitation, and applied framework analysis with double coding to produce a reproducible thematic framework (Connell et al., PLOS ONE, 30 July 2026).
What happened and how the study measured help-seeking
Answer: Connell et al. (PLOS ONE, published 30 July 2026) conducted 50 semi-structured interviews in two Scottish local authority areas between 1 November 2023 and 31 July 2024 to explore influences on help-seeking for mental health and substance use.
Connell et al. recruited via justice social work and third sector services, used Network Canvas to co-create participant sociograms during interviews, and transcribed interviews verbatim before pseudo-anonymisation. The authors then applied framework analysis in NVivo with independent double coding to generate three top-level themes: desire to seek help, ability to seek help, and help-seeking context (Connell et al., PLOS ONE, 30 July 2026).
Direct quotation illustrates the social constraint: Participant P12 said, “I have dealt with it for almost twenty odd years by myself… I don’t like going to doctors because I think they are a waste of time, ” and Participant P13 said, “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, ” both quoted in Connell et al., PLOS ONE (2026).
Findings Snapshot
| Date / Period | Metric | Value | Implication |
|---|---|---|---|
| 1 Nov 2023–31 Jul 2024 | Interviews collected | 50 | Large qualitative sample with rich network diagrams supports a network-focused framework |
| Published 30 Jul 2026 | Participants per area | LA1: 23, LA2: 27 | Contextual contrasts (rural/affluent vs urban/deprived) informed contextual theme differences |
| Data deposit (metadata) | Consent for verbatim quotes | 49 of 50 participants consented | Enables transparent quotation while protecting sensitive data |
| Analysis period | Coding approach | Framework analysis with double coding | Reproducible thematic structure suitable for mapping to interventions |
Implications for qualitative researchers and applied teams
Answer: Connell et al. (PLOS ONE, 30 July 2026) show that qualitative studies of help-seeking must intentionally capture relational data and local context to produce operational findings for intervention design.
Practical implication 1: Elicit social networks during interviews. Connell et al. used participant-created sociograms and Network Canvas during interviews to reveal that network norms and resources determine whether help-seeking will be supported or suppressed.
Practical implication 2: Code for relational processes as well as individual attitudes. The PLOS ONE framework embeds relational sub-themes inside desire and ability rather than treating relationships as a separate, optional category.
Practical implication 3: Design multi-level interventions. Connell et al. recommend testing interventions that act at individual, network, community, and policy levels because networks can share epistemic mistrust and stigma that block individual change.
How Evidano helps qualitative analysis of help-seeking
Problem: eliciting and organizing social network data is time-consuming
Solution: Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Feature: Evidano ingests transcripts and participant-aided sociograms, supports pseudo-anonymisation workflows, and links network metadata to coded excerpts so teams can map relational influence patterns rapidly. See the Evidano features page for details.
Problem: slow manual transcription and inconsistent terminology
Solution: Evidano offers automated transcription with custom dictionaries and PII redaction, reducing turnaround for multi-hour interviews like those in Connell et al. (2026). See the Evidano speech-to-text page.
Feature: Custom dictionaries capture local slang and service names common in criminal justice contexts, improving accuracy for downstream thematic coding.
Problem: synthesising themes across networks and contexts
Solution: Evidano produces thematic, content frequency, and cross-segment analyses that let teams compare themes by network role, locality, or participant subgroup.
Feature: Use Evidano’s AI chat over your documents to query where network mistrust, collective norms, or resource constraints appear across interviews, then export visual co-occurrence networks to inform intervention targets.
Problem: stakeholder-ready outputs and iterative analysis
Solution: Evidano generates shareable visualizations (word clouds, code hierarchies, co-occurrence networks) and supports collaborative review so lived-experience panels and services can inspect findings and co-design responses.
Feature: Secure data handling and the ability to restrict model training make Evidano suitable for sensitive datasets like those described by Connell et al. See data security.
FAQ: qualitative analysis help-seeking criminal justice
How many interviews did Connell et al. conduct and when?
Answer: Connell et al. conducted 50 semi-structured interviews between 1 November 2023 and 31 July 2024, as reported in PLOS ONE (30 July 2026).
Supporting detail: The study recruited 23 people in LA1 and 27 in LA2 and used participant-aided sociograms during interviews to capture social networks (Connell et al., PLOS ONE, 2026).
What is the main transferable finding for qualitative teams?
Answer: The main transferable finding is that help-seeking is relational: network norms, shared mistrust of services, and resource flows shape whether individuals seek and sustain help (Connell et al., PLOS ONE, 30 July 2026).
Supporting detail: The authors explicitly recommend multi-level interventions because relational influences operate across individual and community contexts.
Can AI tools reliably speed analysis of networked qualitative data?
Answer: Yes, when used with researcher oversight, AI-enabled platforms can accelerate transcription, thematic coding, and cross-segment comparisons while preserving analytic rigor.
Supporting detail: Connell et al. used NVivo and framework analysis with double coding for reproducibility; AI tools should support the same checks (double-coding, audit trails) and link to original transcripts for verification.
Is the Connell et al. dataset available for reuse?
Answer: Pseudonymised transcripts and quantitative data are deposited at the UK Data Service under accession 10.5255/UKDA-SN-858261 with access controlled by a data access committee, per Connell et al., PLOS ONE (30 July 2026).
Supporting detail: The paper notes ethical approval and controlled access because the dataset contains sensitive information; contact help@ukdataservice.ac.uk for access procedures.
Conclusion & Next Steps
Answer: The PLOS ONE study by Connell et al. (published 30 July 2026) demonstrates that qualitative analysis of help-seeking in criminal justice-supervised populations must foreground relational and contextual data to produce actionable findings.
For teams designing research or services, prioritize eliciting social networks, use reproducible framework coding, and test multi-level interventions that target network norms and resource flows.
To move from insight to implementation, combine rigorous qualitative workflows with AI-assisted transcription, thematic synthesis, and cross-segment analysis. Try Evidano for free to pilot these workflows in your next study: Try Evidano for free.