From Mixed Methods to Insight: AI Qualitative Analysis

According to the PLOS One study on information technology students, AI qualitative analysis can help mixed-methods researchers connect interview themes with survey results, stakeholder comparisons, and curriculum or policy implications.

The study combined 1, 924 questionnaires with 30 supplementary interviews and proposed a preliminary framework covering general physical fitness, occupation-specific physical fitness, and psychological fitness. The framework remains exploratory because the researchers used convenience sampling, small teacher and enterprise samples, and no structural validity testing.

• In data collected from April 20 to May 29, 2025, 68.71% of the 1, 863 students selected “Frequently” or “Occasionally” when reporting discomfort after professional practical operations.
• The study reported dry eyes or blurred vision among 44.50% of students, cervical or lumbar soreness among 38.06%, and physical fatigue or insufficient endurance among 25.93%.
• The qualitative component included 21 students, 6 teachers, and 3 enterprise personnel, whose interview themes added context about training gaps, teaching resources, and school to enterprise alignment.
• The researchers identified three themes: student occupational physical fitness problems, insufficient institutional resources, and inadequate alignment between enterprise demands and school-based training.
• The study reported that none of the 15 stakeholder comparisons remained statistically supported after false-discovery-rate correction, showing why qualitative context should not be replaced by unadjusted statistical differences.

The 2026 PLOS One study examined how information technology students, teachers, and enterprise personnel perceived occupational physical fitness needs in selected vocational colleges in Shanxi Province, China. The researchers used a quantitative-dominant cross-sectional mixed-methods design, with questionnaires as the primary method and semi-structured interviews as a supplementary explanatory method.

The final analytical sample contained 1, 924 questionnaires: 1, 863 from students, 37 from teachers, and 24 from enterprise personnel. The interview evidence came from 30 participants, including 21 students, 6 teachers, and 3 enterprise personnel.

The authors described the qualitative interviews as supplementary contextual evidence rather than statistical evidence. The researchers stated that the interviews were used to “provide supplementary contextual explanation of questionnaire findings, ” a distinction that matters when integrating qualitative and quantitative results.

The study identified three qualitative themes. Students and teachers described limited occupation-specific training, teachers reported limited access to job-demand information and specialized resources, and enterprise personnel indicated that school-based training did not always reflect workplace requirements.

The researchers organized the proposed framework into three domains: general physical fitness, occupation-specific physical fitness, and psychological fitness. The occupation-specific domain included sedentary tolerance, visual load tolerance, fine motor operation endurance, and static postural endurance.

Mixed-methods researchers can use AI qualitative analysis to make integration more systematic, provided that the analysis preserves source context, sample limitations, and the distinction between themes and measured outcomes.

The first decision is how to connect qualitative evidence to quantitative variables. In the PLOS One study, interview themes about visual relaxation, musculoskeletal relaxation, postural adjustment, and sedentary behavior interruption help contextualize survey findings about discomfort and exercise. A thematic analysis can show what participants mean by a training gap, while a frequency analysis can show how often related responses appear in the survey.

The second decision is how to compare stakeholder groups. The study included students, teachers, and enterprise personnel, but the groups differed substantially in size. AI-assisted cross-segment analysis can organize each group’s language and priorities without implying that a small enterprise sample has the same inferential weight as a large student sample.

The third decision is how to preserve negative or qualifying evidence. The study reported a nominal raw p value for sustained concentration, but the result did not remain supported after false-discovery-rate correction. A responsible AI synthesis should surface that qualification instead of presenting the unadjusted finding as a confirmed group difference.

The study also illustrates why researchers should retain provenance. The authors reported that the original interview audio, verbatim transcripts, field notes, and coding records were no longer available for independent retrospective verification. AI analysis cannot restore missing source material, so researchers should preserve transcripts, codebooks, analytical decisions, and quotations in future projects.

The November 10, 2026 PLOS One study shows how mixed-methods research can combine survey patterns, stakeholder perspectives, and qualitative themes to build a cautious conceptual framework. AI qualitative analysis can make that process faster and more transparent by organizing documents, comparing segments, and surfacing evidence for human review. The strongest workflow keeps source attribution, participant context, statistical corrections, and study limitations visible rather than treating AI-generated themes as final conclusions. Researchers who want to analyze interviews, reports, and survey responses in one workspace can try Evidano for free.