Bridging the Computational Gap in the Social Sciences: Dual Innovations in Pedagogy and Assessment

In a world increasingly dominated by data, graduates in STEM (Science, Technology, Engineering, and Mathematics) fields are in high demand. This focus often obscures the fact that a significant portion of digital data is social in nature, generated through human–computer interactions. Consequently, experts trained in the social sciences should be equally sought after, yet this is not consistently the case. A contributing factor is the tendency for social and political science departments to adhere to traditional disciplinary boundaries, inadvertently marginalizing their graduates from a growing number of data-centric professions. This paper argues that this gap can and should be addressed. We present a pedagogical project initiated at the University of Bologna, centered on a tailored data literacy and social computing course designed for Political Science students. This study evaluates a dual-case pedagogical investigation evaluating two complementary but empirically independent educational initiatives—conducted across distinct courses, student cohorts, and academic years—unified under an overarching conceptual framework for social science data literacy. This framework combines interactive peer-led learning (Case Study 1) with an instructor-driven, backend Generative AI workflow for multiple-choice question generation utilizing Google Gemini via AI Studio (Case Study 2). Crucially, Generative AI tools were utilized strictly as an offline editorial assistant by the instructor to streamline item creation, maintaining a rigorous human-in-the-loop assessment model, while student examinations remained offline and unassisted. The results and student feedback have been noteworthy, demonstrating improved learning outcomes and high satisfaction, despite methodological limitations including sample size constraints and a quasi-experimental design. We share this experience to encourage colleagues in the social and political sciences to replicate and adapt this model, thereby better equipping their students for the demands of the modern professional landscape.