Children as Receptors and Messengers for Health Education – ERMES

The methodology developed in the ERMES project is deeply interdisciplinary; it aims to combine field observations with computer modeling to better understand how children process the dietary messages they receive.

The ERMES project is organized into three WPs.

WP1 (2024–2025 school year) consists of an ethnographic study conducted in four elementary schools (in four fourth-grade classes), two in Paris and two in Seine-Saint-Denis. Participant observation was conducted over the course of a year (one day per week per school) in the classroom, on the playground, and in the cafeteria to identify interactions among peers and with adults. In addition, all food-related messages heard were recorded in a database of arguments.

Three fun workshops gave children the opportunity to express their views on food, engage in discussions, and share their perceptions. Finally, individual interviews with parents at home helped to better understand the students’ social context and gain insight into their family’s eating habits. At the same time, some of the questionnaires (used in WP2) were administered to students in the four classes, and the data obtained complement the data gathered through qualitative observation, serving to either support or contradict the observations made. Finally, the collected data are processed by the project’s computer scientists, who will construct attributed graphs with the students in the observed classes and their characteristics as nodes, and the various types of relationships between the students as edges.

WP2 is a randomized controlled trial involving 298 schoolchildren

Conducted during the 2025–2026 school year, this RCT involves 298 students aged 8 to 10, recruited from ten different schools. The protocol compares two experimental groups (with different message sources) and a control group that receives no messages. Data collection is divided into four phases:

(i) a pre-intervention baseline period (November 2025),

(ii) the intervention period (January 2026),

(iii) an immediate post-intervention phase (February 2026),

(iv) a follow-up conducted in June 2026

The primary outcome measure is the amount of fruits and vegetables consumed in the school cafeteria. Using an AI-based app, consumption is objectively measured by photographing each child’s tray at the beginning and end of each meal, allowing for the assessment of both intake and food waste.

Detailed data on children’s social networks are also collected using a named method, which makes it possible to identify friendships and peer exposure within classrooms.

WP3 aims to disseminate the results of the ERMES project to the scientific community, as well as to practitioners and policymakers. Plans are in place to organize a citizens’ convention on food.

The initial findings from WP1 show how the meanings children attribute to food (its symbolic significance) give it an important role in peer interactions, which are often marked by power dynamics or even domination. Children use food-related messages as tools to address issues of peer group integration or even to gain power or dominance. As for the ERC, the results are not yet available, as data collection is still ongoing.

The ERMES project aims to provide new scientific data to inform the public debate on nutrition education for children in schools. The citizen forum on food, organized as part of WP3, will aim to share the project’s main findings and identify concrete measures to improve the quality of food offered to the public.

The results of the randomized trial conducted in schools offer a unique opportunity to understand children’s eating habits at school, based on evidence-based data and targeted interventions. From a scientific standpoint, this research will be disseminated through the publication of articles in specialized journals, in order to share innovative methodologies and the measured impacts on eating behaviors.

The results will also be presented at conferences and seminars. For public officials, the findings of this study could inform local and national policies on school meals, nutrition education, and the prevention of childhood obesity. Presenting the results to the school and after-school staff involved, as well as to school districts and school boards, could facilitate the incorporation of these recommendations into public health plans that target food.

ERMES is an interdisciplinary research project that mobilizes disciplines belonging to the humanities and social sciences (anthropology, sociology, education sciences), behavioral sciences (decision sciences, psychology, economics) and computer sciences (knowledge representation, machine learning).
By developing a new interdisciplinary method mainly based on the combination of ethnographic, econometric, experimental and modeling approaches, ERMES aims to characterize the mechanisms of transmission of food messages between peers and the effects of peer education programs on eating behaviors. The project will compare such education actions with more commonly deployed approaches to food education (based on teacher intervention). Moreover ERMES proposes to explore the food messages that children receive and relay (especially those related to their health or environmental consequences) and to better understand which profiles of children are most likely to be influenced by these messages.
The study will take place in 14 elementary schools (24 fourth grade classes) in the city of Paris (socially and economically contrasted) and the department of Seine-Saint-Denis (socially and economically contrasted).
WP1 will consist of an ethnographic study in four fourth grade classes (complemented by quantitative data from questionnaires that will be used in WP2) to better understand the underlying mechanisms, in particular the role of group dynamics in children's dietary behaviors and discourses. WP2 proposes to study the short and medium term effects of food messages issued either by peers or by teachers on the knowledge, attitudes and behaviors of the recipient children, with a randomized controlled trial conducted in 500 pupils. WP3 aims at presenting the results of WP1 and WP2 to different circles of experts and civil society actors in order to formulate proposals for guidelines for public action in the field of food education. The work carried out in this WP3 will also contribute more broadly to the study of behavior change and to public action on health-related behavior change.
The originality of the project lies in the study of the role and influence of friendship networks on the circulation of food messages among children and their consequences on food discourse and behavior. It will be based on a mapping of the groups of friends in each class, notably by means of a questionnaire that has never been used in France. ERMES comes from a multi-actor research project supported since 2020 by the ARS Île-de-France, bringing together scientists, a local authority in Seine-Saint-Denis, school and extracurricular professionals, as well as associative actors who deal with health promotion among young people (MODALITEA project). This framework will ensure effective dissemination of research results and, in particular, improve health education programs and specify the conditions that will make these programs more effective.