The work addresses the significance of robotics in education, emphasizing its role in enhancing STEM skills through programming and sensory feedback. Scratch, a multimedia programming environment, is highlighted as a tool for robotic projects. Within Scratch, this work discusses data representation, distinguishing between visible and transparent data, focusing on transparent data. To evaluate how data are used in Scratch by novice programmers, this work also proposes an evaluation framework using the Structure of the Observed Learning Outcome (SOLO) taxonomy.

Robotics is being increasingly used in order to engage pupils in STEM, in which computer programing can also be taught. This paper, as part of a larger research approach within Robotics and STEM where visual programming languages are used and proposes a set of evaluation criteria on data assessment in Scratch based on the SOLO taxonomy. The data evaluation is influenced by their form, the types and symbolism of their content, the ways in which they are used, their actors, their degree of visibility, the organization of simple and multimedia data, their processing, and their naming. The types and symbolism of their content, as well as the ways in which they are used, are investigated in this work. For each considered factor, a categorization is made using the SOLO taxonomy levels. The teachers can use this categorization as a tool to either assess their students' performance using quantifiable criteria (in matters pertaining to the factors mentioned above) or to create their own teaching paths, thereby building a mental scaffolding that aids in the efficient and effective learning of new material.

According to Wirth, “a program is algorithm plus data”. This association of algorithm and data also occurs at the micro-scale of a command, where the program is represented by a single command plus a data representation. This paper proposes to map the degrees of complexity of the command, data processing and data structuring to the levels of the SOLO taxonomy in order to create tool for evaluating code which is used in visual programming environments. This tool allows the teacher to assess the degree of the student’s understanding of the use of a command, and the complexity of that command. The ability of the teacher to measure the weight of each command, allows him to design alternative teaching spiral paths, from a simple command to the most complex ones. The work in this paper is part of a broader project to evaluate visual programming code – with Scratch as the chosen visual programming language.

This paper aims to investigate, classify and establish appropriate benchmarks for visual programming codes. For the present study we have relied on the use of the Scratch-2 visual programming language. Scratch-2 is very popular as an introductory programming language at primary education and at the first years of secondary education in Greece. Initially, an evaluation framework, named VPLEF (Visual Programming Languages Evaluation Framework), is constructed based on several selected criteria. Since the observation of the visual code is not easy in the programming environment, we adopted the codOrama tool for a more consistent representation of the Sctrach-2 codes, aiming to identify the code characteristics, which were then associated with the evaluation criteria. We focuson the evaluation criteria of user interactivity, communication and parallelism-linearity of VPLEF and on two groups of commands, namely on “Control” and Event”, in conjunction with the system arguments. The study of the aforementioned entities leads us to a set of specified relationships of each of these three criteria with the command/argument combinations. Finally, an evaluation process is proposed to quantify the contribution of each combination to a specific criterion.