Student Insights: AI Qualitative Analysis of School Food

According to the November 3, 2026 PLoS One research article, secondary school students in Northern Benin described school food environments as widely available but often unhealthy, unhygienic, expensive, and limited in fruits and vegetables.

The study shows why qualitative analysis of focus groups should examine perceived access, safety, affordability, culture, and urban-rural differences rather than counting food outlets alone.

• The PLoS One study collected data from 40 students in four focus group discussions across Natitingou, an urban municipality, and Boukombe, a rural municipality, during November and December 2022.
• The PLoS One researchers reported that available foods were predominantly energy-dense or starch-based, while fruits and vegetables were less accessible because of seasonality, cost, and limited physical access.
• The PLoS One study identified 10 fruits and 18 vegetables mentioned as available in or around the schools, but students reported consuming fruit approximately two to three times per week.
• The PLoS One researchers found that urban students perceived greater food variety, while rural students reported sharper seasonal constraints and no outlet offering vegetable-based dishes within the 1-kilometer study radius.
• The PLoS One authors proposed student-informed actions including nutrition education, vendor hygiene training, school gardens and orchards, food preservation, and support for local production.

The November 3, 2026 PLoS One study examined how secondary school students perceived food availability, preferences, affordability, hygiene, and possible improvements in two municipalities in Northern Benin.

The researchers held two focus group discussions in each municipality, separating lower-secondary students aged 11 to 15 from upper-secondary students aged 14 to 19 and maintaining equal gender representation. Each discussion lasted approximately 45 to 60 minutes and covered food available inside schools and within 1 kilometer of school premises.

The PLoS One research team conducted discussions in French and local languages, including Ditammari and Waama. The team transcribed recordings, translated relevant passages, used back-translation, and applied a hybrid deductive-inductive coding approach.

Two researchers independently coded a subset of transcripts with MAXQDA version 24, discussed discrepancies, refined the codebook, and completed comparative analysis between urban and rural settings. The study reported no meaningful differences in themes between school levels.

A student in Boukombe summarized the perceived gap between availability and quality by saying, "Our environment is very, very poor in terms of food in general." The quotation comes from the November 3, 2026 PLoS One study and represents one participant’s perception, not an objective measurement of every school food setting.

The central implication from the November 3, 2026 PLoS One study is that student perceptions add information that objective food audits may miss.

The PLoS One researchers distinguished between what is physically available and what students perceive as affordable, safe, desirable, and culturally acceptable. Students described high prices, poor hygiene, seasonal shortages, religious or family restrictions, and concerns about food preparation as factors shaping choices.

The study also shows why segmentation matters. According to the PLoS One findings, Natitingou students perceived greater food variety, while Boukombe students reported stronger dependence on market days, agricultural cycles, and limited local supply. Shared concerns across both settings included affordability, hygiene, and the dominance of starch-based foods.

For health researchers, the practical workflow is to code both the barrier and the mechanism: for example, seasonal scarcity can lead to reduced access, high prices can reduce purchase likelihood, and vendor hygiene concerns can reduce trust. AI-assisted thematic analysis can surface these relationships quickly, but researchers should validate interpretations against transcripts, translations, and participant quotations.

The study’s recommendations should be treated as policy implications rather than proven intervention effects. The authors explicitly noted that the qualitative design could identify perceived barriers and proposed solutions, but could not establish causal relationships or demonstrate that an intervention would improve diets.

The November 3, 2026 PLoS One study demonstrates how student-centered qualitative research can reveal the gap between food being present and food being trusted, affordable, nutritious, and culturally acceptable.

AI qualitative analysis can help health researchers process multilingual transcripts, identify recurring themes, compare segments, retrieve evidence, and turn focus group data into a transparent research workflow.

Researchers should use AI as an auditable assistant, retain human review, protect participant confidentiality, and avoid treating qualitative frequencies as population estimates.

Teams analyzing interviews, focus groups, open-ended surveys, and policy documents can try Evidano for free to accelerate qualitative synthesis while keeping the research context visible.