Computational thinking, a concept famously articulated by Jeannette Wing in 2006, refers to the ability to formulate and solve problems using models drawn from computer science. It encompasses abstraction, decomposition, generalisation, algorithmic thinking and evaluation, and it has been repeatedly linked to stronger problem-solving, pattern recognition and information selection. Because educational authorities now regard it as an essential competence, teacher educators face a pressing question: how do you cultivate computational thinking in people who will teach three- to six-year-olds, many of whom arrive at university with little or no programming background? The Spanish team behind StoryMat-Robot believed the answer might lie not in code alone, but in stories.

The premise of the project is deceptively simple. A StoryMat-Robot is a playful, interactive narrative starring a robot, which must move through a physical space, typically a mat decorated with settings, characters and obstacles drawn from an adapted story. The narrative must be coherently sequenced, synchronising the plot’s progression with the robot’s trajectory, while the robot overcomes playful challenges tied to the storyline. Participants in the study worked in groups of four or five to produce 38 such proposals, adapting works ranging from Hansel and Gretel, The Three Little Pigs and The Wizard of Oz to animated films such as Ratatouille, Up and Madagascar. Roughly 60 percent of the designs were based on children’s literary texts and 40 percent on films.

What makes the design process cognitively rich is the way it forces several distinct skills to operate simultaneously. Translating a plot into a visual, physical format demands abstraction: students must select the most relevant story elements, represent them graphically as a robot pathway, and adapt characters while preserving their defining traits. Breaking the storyline into narrative sequences, scalable challenges and structured robot movements exercises decomposition. Maintaining aesthetic and stylistic coherence with the original work, whether two-dimensional, three-dimensional or realistic, mirrors generalisation, and segmenting the story into curriculum-aligned educational challenges with clear objectives, feedback and timing draws directly on algorithmic thinking. Layered over all of this are transmedia competences, as students integrate digital resources, augmented reality applications and artificial intelligence tools into their mats.

To evaluate what the finished products revealed about their creators’ thinking, the researchers designed and validated a 24-indicator instrument organised into four dimensions and scored on a four-point rubric, from ‘not at all adequate’ to ‘very adequate’. Three independent raters, including an external researcher, assessed every StoryMat-Robot, achieving strong inter-rater reliability with an intraclass correlation coefficient of 0.874. An exploratory factor analysis confirmed the instrument’s validity, with three factors explaining 64.9 percent of the variance and a high internal consistency of alpha equal to 0.946. Statistical comparisons between literary-based and film-based designs employed the Mann-Whitney U test, complemented by effect sizes and confidence intervals.

The findings were encouraging across the board. Overall adequacy reached moderate-to-high levels, with decomposition scoring highest at 3.36, followed closely by algorithmic thinking at 3.34 and abstraction at 3.26, while generalisation lagged slightly at 2.98. The strongest single feature was the incorporation of educational challenges appropriate for early childhood education, rated at 3.69, suggesting that the trainee teachers excelled at embedding pattern-recognition tasks and age-appropriate problems into their narratives. In one adaptation of The Three Little Pigs, children were asked to select building materials; in the Ratatouille mat, they collected ingredients for the traditional dish. Students also showed a strong ability to adapt characters to the interactive format while preserving the essence of the original stories.

Interesting differences emerged between the two source materials. Designs based on literary texts proved more robust in spatial organisation: the logic of the robot’s pathways, the scaling of challenges, the alignment of tasks with the educational level and the overall mat design all scored significantly higher, likely because short, simple stories such as Hansel and Gretel often contain explicit routes that students could transfer directly onto the mat. Film-based designs, by contrast, were richer in digital integration, scoring higher on the use of artificial intelligence resources, the creation of augmented reality elements aligned with the narrative, the relevance of educational objectives, the functionality of digital tools and the inclusion of assessment procedures. The researchers attribute this to the greater multimodality of audiovisual discourse, which seems to prime designers toward technological enrichment and stronger pedagogical framing.

Not every dimension flourished equally. Generalisation proved the weakest component, and the integration of coding cards explaining the robot’s movement sequences was rated particularly low at 1.90 overall, revealing that trainee teachers struggle to make programming patterns explicit and transferable. The researchers also caution that the statistically significant differences between literary and film-based designs carried small or even negligible effect sizes, meaning neither format is categorically superior. Rather, the near-uniform presence of computational thinking skills across both groups suggests that the co-design process itself, grounded in Design-Based Learning, is the active ingredient. The study’s limitations are acknowledged candidly: it took place in a single institutional context, computational thinking was inferred from products rather than measured directly with pre- and post-tests, and students chose their own source works, a potential confounding factor.

The implications reach beyond one teacher-education classroom. The study reinforces a growing body of evidence that narrative creation, digital storytelling and educational robotics are powerful vehicles for computational thinking, and it demonstrates that these can be fused into a single, pedagogically meaningful activity rather than treated as separate strands. Because the resulting mats are ready-made classroom resources, the benefits flow in two directions: future teachers practise abstraction, decomposition and algorithmic sequencing, while the young children who eventually use the mats encounter the same skills through play. The research team now proposes testing the StoryMat-Robots with actual early childhood pupils aged three to six, to measure their influence on learning, motivation and computational thinking. If those trials succeed, the humble story mat may become a standard fixture in the effort to raise a generation that thinks computationally before it can even write its own name.

Subject of Research: Developing computational thinking in prospective early childhood teachers through robot-based transmedia storytelling design

Article Title: StoryMat-Robot: From Activating Narrative and Transmedia Competences in Prospective Teachers to the Development of Computational Thinking

Article References: StoryMat-Robot: From Activating Narrative and Transmedia Competences in Prospective Teachers to the Development of Computational Thinking. (n.d.). https://doi.org/10.1007/s13158-026-00544-7

Image Credits: AI Generated

DOI: 10.1007/s13158-026-00544-7

Keywords: computational thinking, educational robotics, transmedia storytelling, teacher education, early childhood education, narrative competence, augmented reality, artificial intelligence, algorithmic thinking, abstraction, decomposition, design-based learning