Learning data storytelling and SPARQL in the humanities: Tracking the co-evolution of skills

This study investigates how humanities students develop SPARQL querying skills, data visualisation literacy, and data storytelling quality using MELODY, a tool for authoring Linked Open Data stories. We analysed 51 student data stories, examining textual, structural, and query features and explored correlations between technical proficiency, narrative coherence, readability, alternation, and visualisation practices. Results reveal a complex picture: SPARQL proficiency aligns moderately with narrative coherence, while alternation, readability, and information density evolve independently. Collaboration shows limited impact on skills, and student sentiment highlights challenges with incomplete datasets and tool usability, alongside appreciation for visualisations that support iterative query refinement. These findings suggest that students prioritise either technical accuracy or narrative clarity, rarely excelling at both simultaneously. Implications for teaching include scaffolding technical and storytelling skills together, leveraging visualisations for iterative learning, and structuring collaboration effectively.