The integration of computational thinking (CT) into mathematics education has attracted increasing attention; however, there is limited methodological clarity about how CT can be systematically embedded in primary mathematics instruction. In particular, empirical research linking CT to multi-step mathematical problem-solving remains scarce. This paper presents the design and implementation of a task-based intervention model developed within the DigiMaths4All project. The model incorporates CT into primary mathematics through a structured framework for task selection and design, combined with implementation in a technology-enhanced learning environment (ViLLE). Mathematical tasks focus on arithmetic fluency and multi-step problem solving, while CT tasks are tailored to operationalize key processes such as decomposition, abstraction, algorithmic thinking, and debugging. The intervention was carried out through a class-based randomized controlled trial in primary education, comparing traditional instruction with technology-supported approaches, including an integrated mathematics CT model. The study uses a mixed-methods approach, incorporating assessments and learning analytics data to examine implementation processes. The main contribution of this paper is methodological. It offers a replicable framework for (1) designing interventions that incorporate CT into mathematics, (2) selecting and constructing aligned task sets, and (3) implementing these tasks within an analytics-driven digital environment. The findings enhance understanding of how CT can be operationalized to support mathematical problem solving in primary education.
This study addresses an implementation problem for informatics education: whether teachers’ reported participation in AI-related professional learning, interpreted as realised access to one professional learning condition for teacher AI literacy, is associated with declared need or instead follows existing patterns of digital, professional and organisational advantage. The study does not measure teacher AI literacy, AI competence, computing teaching practice, computational-thinking instruction or classroom implementation directly. Rather, it analyses reported participation in AI-related professional learning as a realised opportunity condition for developing teacher AI literacy at scale. Using TALIS 2024 data from 108,136 lower-secondary teachers nested within 10,840 schools across 55 education systems, three-level multilevel linear probability models and random slopes at the education-system level were estimated. Results showed substantial cross-system inequality in reported participation. Variance decomposition located 8.5% of total variation at the education-system level and 9.7% at the school level. Declared need was only partially associated with reported participation: teachers reporting low or moderate need were more likely to have participated in AI-related professional learning than those reporting no need, whereas teachers with the highest need showed no significant advantage. Digital self-efficacy and professional collaboration were consistently associated with higher participation. At the school level, digital resource shortages and school digital leadership support were significant predictors. Random-slope estimates showed that the association between high declared need and participation varied significantly across education systems. The findings suggest that equitable teacher AI literacy requires deliberate opportunity structures, not only competence frameworks or voluntary participation in professional learning.
Debugging is integral to programming. It comes into play as soon as novices make their first mistakes in creating programming artifacts. It is also consistently reported to be a skill that is difficult to learn as well as to teach effectively. Research in Informatics Education has often focused on the process of debugging, by breaking it down in steps connected by temporal and causal dependencies. In this work, we focus instead on debugging as a skill, from the standpoint of Cognitive Load Theory, and break it down into a tree-shaped model of subskills that enable one another. Debugging may thus be seen as a meta-skill that requires the coordination of multiple others. From the standpoint of Cognitive Load Theory, such a skill is cognitively expensive, which may explain the learning-related difficulties tied to debugging. Using the framework of the four-component instructional design, we hypothesize a categorization of each debugging subskill as either recurrent or nonrecurrent, dividing those that are applied consistently to different contexts from those that require problem solving. All subskills may be practised and potentially assessed with targeted exercises, whose design depends on their recurrent/nonrecurrent nature. We provide extensive examples of such exercises. Our decomposition of debugging into subskills is a novel way to address debugging in educational contexts and complements the work done on debugging processes. Although it is currently a theoretically grounded conjecture, the model provides concrete guidance for instructors on analyzing existing materials and planning cognitive-load-informed learning trajectories.
Source code comments are usually treated in software engineering as documentation artifacts that support readability, maintainability, and long-term comprehension. In programming education, however, comments may also function as pedagogical scaffolds by helping learners externalize reasoning, clarify intent, and reflect on code. This article presents a PRISMA-informed qualitative systematic review with layered evidence on the pedagogical role of source code comments in programming education. Searches in Scopus and ERIC produced 50 unique records; 36 were assessed in detail and resolved into 18 primary synthesis studies, 7 supporting/contextual studies, and 11 advanced-stage exclusions. Because the evidence base is heterogeneous, the review uses qualitative layered synthesis rather than meta-analysis. The findings suggest that comments are best understood pedagogically as explanation-centered learning supports, especially for code comprehension, self-explanation, debugging and reflective reasoning, and formative assessment of student thinking. Direct comment-focused evidence remains limited and concentrated mainly in highereducation and novice programming contexts; adjacent explanation-centered studies clarify plausible mechanisms but do not by themselves establish comment-specific effects. The review concludes that comments can become pedagogically meaningful when deliberately integrated as scaffolds for explanation, comprehension, and reflection.
We investigate the pedagogical impact of Graphical Loop Invariant Based Programming (GLIBP) in an introductory programming course. This approach encourages students to visually model the objects and variables handled in the loop, before implementing it. To evaluate the efficiency of this GLI model, a four-condition A/B/C/D test was conducted across two problems, with students receiving varying levels of scaffolding (from no support to a fully constructed GLI). Analysis of students’ code showed that a well-designed GLI reduced errors related to the loop guard and the update of variables. However, many students struggled to understand or represent a GLI. The fill-in-the-blank GLI version, in particular, often added cognitive load rather than reducing it. Three recommendations emerged: train students to interpret a provided GLI when writing code; second, teach students to sketch their own model by recognizing similarities to previously solved problems; finally, guide students with questions to ensure all necessary variables and relationships are properly identified.
