This presentation highlights a problem-based statistical literacy model developed within the academic extension project "Learning and Relearning with Citizen Data from Real Problems" at ENCE/IBGE (the National School of Statistical Sciences of the Brazilian Institute of Geography and Statistics). Combining statistics, communication, and social sciences, this interdisciplinary framework pairs future statisticians with Civil Society Organizations (CSOs) to tackle complex, real-world community challenges. In 2026, we had two different approaches: in the first case the problem arose on the first visit to the CSO, and in the second case a different CSO brought an unstructured problem. In the first approach, where the central issue was initially unknown, students engaged in field research, reflective workshops, and conceptual mapping aligned with the Copenhagen Framework and data value chain. Moving beyond an initial surface assumption ("lack of funding"), students identified the root cause as a data governance deficit in tracking assisted beneficiaries. This shift demonstrated how structural data management directly dictates organizational efficiency and public transparency. In the second approach, students addressed an unstructured problem: designing measurement instruments to evaluate an CSO’s impact on women entrepreneurs. Through iterative interactions with the CSO, the team noticed that assessing social impact required exploring complex sample comparability challenges between treatment and control groups. Through multiple iterations, students navigated the full data value chain—from conceptual design and sample framing to field collection and contextual interpretation. These hands-on trajectories bridged abstract statistical theory with the complex realities of social impact, building capacities for independent data production in OSCs while expanding the social consciousness, critical thinking, and ethical responsibility of future statistical professionals.
