
Final-year research project in urban engineering at UTC · from literature review to a school classroom
Science communication is usually treated as something downstream of research: the results are in, then we figure out how to explain them to the public. This TZ asks about the other side: if science mediation happens upstream, can it be the method itself?
The report is titled “Développement méthodologique par la médiation scientifique” (methodological development through science mediation), supervised by Maxime Hachette and Manuela Sechilariu. It is not a portfolio of outreach pieces but a methodological study: literature, field, proposal.
What the close reading yields is not a catalogue of definitions but four tensions:
On 1 April 2026, at Lycée Charles de Gaulle in Compiègne, I assisted — as an assistant observer — a postdoctoral researcher from the AVENUES lab during a science-mediation session on electric mobility.
My position was not that of the main speaker but of an observer placed close to the device: preparing the material, following the session, watching students’ reactions, noting the moments of attention, understanding and hesitation.
Three observations:
This observation pulls mediation from the level of texts back to the level of the device: a mediation session can surface clues about understanding, representations and cognitive obstacles.
Table 1 in the report turns the mediation setting into research material:
| Device | What is observed | Data produced | Potential use for research |
|---|---|---|---|
| Interactive quiz | Correct or wrong answers, hesitations, level | Indicators of understanding, fuzzy notions, representations | Building cognitive and informational variables |
| Participatory workshop | Exchanges, arguments, prioritisation of problems | Verbatims, perceived obstacles, usage categories | Enriching hypotheses, discussing indicators |
| Visual support / slides / infographics | Attention, appropriation, ability to reformulate | Indicators of readability and clarity | Testing the usability of research outputs |
| Discussion with actors | Operational constraints, disagreements, validation | Experience feedback, institutional needs, context variables | Adjusting the operational relevance of tools |
Hence a four-step framework:
Mediation is no longer just “knowledge transfer” at the end of research, but a space where data emerge: it makes visible dimensions that spatial or technical data alone struggle to capture — the legibility of devices, cognitive load, informational obstacles, and the gap between theoretical accessibility and actual appropriation.
The report is explicit about its limits: the review is targeted rather than exhaustive, the close-read sub-corpus is small, and the framework is a proposal, not a protocol already implemented at scale.
This TZ grew directly into a doctoral application: the SOUND project builds science mediation into the research design as a methodological component (32 days per year).

