Primary keyword: AI qualitative analysis mixed-methods. According to the PLOS ONE protocol by Phonyiam et al. (published July 29, 2026), the study uses an explanatory sequential mixed-methods design to examine social determinants of health and postnatal well-being among Thai women with type 2 diabetes. Researchers and qualitative teams can use AI qualitative analysis mixed-methods techniques to speed translation, codebook development, thematic synthesis, and joint-display integration while preserving traceability and participant privacy.
Key Takeaways
According to the PLOS ONE protocol by Phonyiam et al. (published July 29, 2026), an explanatory sequential mixed-methods study will survey 50 postnatal women with type 2 diabetes and interview 12 purposively sampled participants to link social determinants of health to postnatal well-being. PLOS ONE
- 50 participants will complete the quantitative survey, and 12 participants will be interviewed, recruitment began in September 2025 and is expected to finish by December 2026, with results projected by June 2027.
- The Thai translation of the Postnatal Well-being in Transition instrument achieved I-CVI and S-CVI values of 1.0 in the expert validation, as reported in July 2026.
- The study uses a three-step mixed-methods integration (fit assessment, procedural resolution, synthesis) to produce meta-inferences, following Fetters et al., 2013, as cited in PLOS ONE (Phonyiam et al., 2026).
What happened and how the study works
What happened: Phonyiam et al. published a study protocol in PLOS ONE on July 29, 2026 that lays out an explanatory sequential mixed-methods study in two public hospitals in Thailand.
How it works: According to Phonyiam et al. (PLOS ONE, 2026), Phase 1 completed a six-stage cultural translation and validation of the Postnatal Well-being in Transition scale, and Phase 2 will collect quantitative data from 50 postnatal women and qualitative interviews with 12 nested participants selected from the survey extremes.
Measurement and timing: Phonyiam et al. (PLOS ONE, 2026) describe a 30-item postnatal well-being Likert instrument, a 20-item SDoH frequency tool, SPSS analysis for means and frequencies, Atlas.ti v9 for coding, and joint-display triangulation to integrate qualitative themes with quantitative means.
Constraints: Phonyiam et al. (PLOS ONE, 2026) note a limitation that the adapted SDoH instrument lacks formal cross-cultural psychometric validation for Thai populations, so findings will require cautious interpretation for other Southeast Asian contexts.
Findings snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| July 29, 2026 | Protocol published | PLOS ONE article (Phonyiam et al.) | Public protocol enables pre-registered mixed-methods procedures and transparency |
| September 2025 | Recruitment started | Quantitative and qualitative recruitment (urban + suburban sites) | Enables maximum variation sampling across socioeconomic contexts |
| Target period Sep 2025–Dec 2026 | Sample sizes | n=50 (survey), n=12 (interviews) | Descriptive quantitative trends with nested qualitative explanation |
| July 2026 | Content validity | I-CVI = 1.0, S-CVI/Ave = 1.0 | Expert panel rated Thai translation as highly relevant before field testing |
Implications for nursing researchers and qualitative teams
Implication summary: Phonyiam et al. (PLOS ONE, 2026) show that explanatory sequential mixed-methods yields both population-level patterns and contextual explanations, which is useful when social determinants interact with chronic disease transitions such as postpartum T2DM.
Design decisions: According to Phonyiam et al. (PLOS ONE, 2026), researchers should plan nested sampling to re-contact extreme quantitative cases (6 lowest and 6 highest well-being scores) to maximize explanatory depth and to code blind to scores during initial qualitative analysis to reduce bias.
Measurement guidance: Phonyiam et al. (PLOS ONE, 2026) recommend side-by-side joint displays and a three-step interpretive framework (fit assessment, procedural resolution of discordance, synthesis into meta-inferences) to reconcile convergence, expansion, and discordance between datasets.
How Evidano helps: mapping study needs to AI-enabled features
Problem: Slow translation and instrument adaptation → Solution: faster, auditable translation workflows
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
How Evidano maps: the PLOS ONE protocol (Phonyiam et al., 2026) required a six-stage translation and back-translation process; Evidano’s translation workflows and custom dictionary support can accelerate forward/back translation and produce a version-controlled trail for ethics boards (Features, Translation).
Problem: Manual transcription, privacy overhead → Solution: secure transcription and redaction
How Evidano maps: Phonyiam et al. (PLOS ONE, 2026) plan verbatim Thai audio transcription and de-identification; Evidano’s speech-to-text with custom dictionaries and PII redaction reduces manual work while preserving audit logs (Speech-to-Text).
Problem: Integrating quantitative means with qualitative themes → Solution: AI-assisted joint displays and thematic synthesis
How Evidano maps: Phonyiam et al. (PLOS ONE, 2026) emphasize side-by-side joint displays and a three-step synthesis; Evidano generates thematic, frequency, and cross-segment analyses and exports crosstab-style joint displays to speed the meta-inference step.
FAQ: AI qualitative analysis mixed-methods
How can AI speed explanatory sequential mixed-methods analysis?
Direct answer: AI speeds routine tasks so researchers can focus on interpretation.
Supporting detail: According to Phonyiam et al. (PLOS ONE, 2026), explanatory sequential designs require iterative translation, survey scoring, and qualitative coding; AI can automate transcription, suggest initial codes, and produce statistics-by-themes joint displays for rapid fit assessment.
Is automated transcription acceptable for Thai-language interviews?
Direct answer: Automated transcription is acceptable when combined with human review and custom dictionaries.
Supporting detail: Phonyiam et al. (PLOS ONE, 2026) plan verbatim Thai transcription with a research assistant review; Evidano’s speech-to-text supports custom dictionaries and human-in-the-loop correction to match that protocol (Speech-to-Text).
How does AI assist mixed-methods triangulation and joint displays?
Direct answer: AI can auto-populate joint displays and flag areas of convergence, expansion, or discordance for human review.
Supporting detail: Phonyiam et al. (PLOS ONE, 2026) use a three-step framework for integration; AI tools can compute means and SDs, link quotes to score strata, and generate side-by-side tables for researcher-led synthesis.
Will AI tools preserve participant privacy and research ethics?
Direct answer: Yes, when the platform enforces encryption and PII redaction and keeps user data isolated from third-party training sets.
Supporting detail: Phonyiam et al. (PLOS ONE, 2026) describe secure storage and de-identification for three years; teams should choose platforms that offer encrypted storage, access controls, and documented retention policies.
Conclusion & Next Steps
Phonyiam et al. (PLOS ONE, published July 29, 2026) provide a transparent protocol showing how an explanatory sequential mixed-methods design can illuminate the social determinants of postnatal well-being for women with type 2 diabetes.
Researchers should plan for nested sampling, validated translations, and a documented integration strategy as described in the protocol.
If your team needs reproducible transcription, translation, thematic synthesis, and joint-display export for an explanatory sequential study, consider tools that integrate these steps to cut analysis time and strengthen audit trails.
To evaluate these workflows in your project, Try Evidano for free.