We study how students and teachers conceptualize the notion of 'algorithm', a fundamental concept in computer science and computer science curricula. We analyze the work produced by CS students and teachers during a workshop conducted repeatedly over several years in some outreach activities for schools, computing education courses, and professional development opportunities for teachers. Participants were divided into groups, given some procedures written in natural language, and asked to decide together which of the procedures might be taken as algorithms. The procedures were purposely designed to present flaws or features that could activate discussion. After that, groups were asked to agree upon a definition of 'algorithm' and make its fundamental properties explicit. The activity triggered reflections around the idea of algorithm and its interpreter, going beyond stereotyped definitions, and leading participants to deepen their comprehension of the notion. We report on the aspects that were more debated by the groups, and those that were recurrent in the resulting definitions. We argue that this kind of activities should be offered more often both to students during their study career, and to teachers in professional development opportunities, to prompt them to reflect on computing foundations also in a non-technical, more holistic way.

In this paper we study the problem-solving ability of the Large Language Model known as GPT-3 (codename DaVinci), by considering its performance in solving tasks proposed in the “Bebras International Challenge on Informatics and Computational Thinking”. In our experiment, GPT-3 was able to answer with a majority of correct answers about one third of the Bebras tasks we submitted to it. The linguistic fluency of GPT-3 is impressive and, at a first reading, its explanations sound coherent, on-topic and authoritative; however the answers it produced are in fact erratic and the explanations often questionable or plainly wrong. The tasks in which the system performs better are those that describe a procedure, asking to execute it on a specific instance of the problem. Tasks solvable with simple, one-step deductive reasoning are more likely to obtain better answers and explanations. Synthesis tasks, or tasks that require a more complex logical consistency get the most incorrect answers.

In a project partially supported by research grant PANN20_00690 to Italy's CINI National Lab "Informatica e Scuola", we compared the effectiveness of two alternative instructional methods applied to scaffold the learning of iterations for children at grades 2-3. Eight university groups across the Country collaboratively run the project in two successive rounds throughout the year 2022. Teachers' feedback collected across the two rounds helped fine-tune the deployment of the interventions. The experiment results show that the two alternative interventions have measurable outcome differences in the short term.

This Research Full Paper presents a study of programming strategies to manipulate data series presented in a range of CS1 courses. We collected and analyzed programs from multiple institutions in different countries, covering a range of programming languages (Python, Java, Go, and C). We started from a list of core strategies (that is, strategies that process the series as a whole, e.g., goals/plans for counting, linear search, etc.) drawn from the literature. We then expanded and refined the list, based on the analysis of the collected programs. We used a mixed method: we first analyzed the programs qualitatively to identify the recurring goals; then we conducted a quantitative analysis of the frequencies of goals. The result of the qualitative analysis is a data-driven list of basic goals that are required in order to cover the input, storage, and processing of data series, independently of the chosen programming language. The list has 13 well-separated goals (no goal in the list is a sub-goal of another one) that share a similar structure and the same level of abstraction. The quantitative analysis shows that the category of core goals is, not surprisingly, the most frequent one; around half of the programs contain a core goal, with total, count, and linear search being the most recurrent ones. Besides them, goals related to input and storage play a significant role as well: taken together, they occur with a frequency similar to that of core goals. The list of goals, if used to select or design practical exercises and teaching materials, can be a valid reference for CS1 instructors who want to foster the acquisition of strategic knowledge by their students. Implication for teaching and learning are discussed.

In order to support students struggling with learning to program, we designed new learning material, around two basic principles from recent research on introductory programming education: propose program comprehension tasks[4] beside writing tasks, and teaching strategic programming knowledge explicitly[5].

When learning to program, different types of knowledge and skills have to be built: syntactic knowledge, i.e., knowledge of the language features; conceptual knowledge, i.e., knowledge of the semantics of the various constructs; strategic knowledge, i.e. the ability to use syntactic and conceptual knowledge in the most appropriate and effective way to solve programming problems. Strategic knowledge also includes the ability to exploit and combine solutions to known problems to solve new ones and is one of the hardest skills to acquire when learning to programm [5, 6, 11]. Terms such as goals[10] patterns [5, 6, 11]. schema,[9] strategies[4], and variables' roles[3] have been used to refer to this kind of knowledge. Several authors suggest that strategic knowledge should be taught explicitly, however there are three main issues to address:

Program-writing skills are often assessed with computer-based practical exams where students are required to design, write and test small programs according to specifications written in natural language. The evaluation of such tasks is usually based on the correctness of outputs (possibly with the support of automatic tools) and the quality of code. In this paper, we propose to evaluate program-writing assignments by using Soloway's notion of plans [34]. We illustrate this idea by defining a marking scheme for some typical program-writing tasks and apply this approach to evaluate the programs written during the final examination of an introductory programming course. We then compare the resulting marks with the marks originally assigned by the instructors of the course. Findings show that the use of plans in the marking process is feasible and requires an effort similar to the one required by a traditional evaluation process, and the focus on plans when examining the source code of students gives useful insights about their difficulties.

The Bebras International Challenge on Informatics and Computational Thinking is a contest open to pupils of all school levels (from primary up to upper secondary) based on tasks rooted on core informatics concepts, yet independent of specific previous knowledge such as for instance that acquired during curricular activities.This paper describes the design choices, the architecture, and the main features of the web-based platform used to carry out the Italian Bebras contest.This platform includes functionalities needed by students, teachers, and Bebras staff during the execution of the challenge, tools to support the preparation of tasks and the training of students, instruments to evaluate the results and analyse data collected during the challenge.The platform is online since 2015 and it has managed the participation of around 25,000 teams and a significant amount of training sessions.

We present a new classification method for Bebras tasks based on the ISTE/CSTA operational definition of computational thinking. The classification can be appreciated by teachers without a formal education in informatics and it helps in detecting the cognitive skills involved by tasks, and makes their educational potential more explicit.

This paper analyses the results of the 2014 edition of the Italian Bebras/Kangourou contest, exploiting the Item Response Theory statistical methodology in order to infer the difficulty of each of the proposed tasks starting from the scores attained by the participants. Such kind of analysis, enabling the organizers of the contest to check whether or not the difficulty perceived by pupils was substantially different from that estimated by those who proposed the tasks, is important as a feedback in order to gain knowledge to be used both in ranking participants and in organizing future editions of the contest. We show how the proposed analysis essentially highlights that the 63% of tasks was perceived at the same level of difficulty estimated by those who proposed them, but a 37% of tasks were either easier or more difficult than expected.

This article describes the state of informatics education in the Italian secondary schools, highlighting how the learning objectives set up by the Ministry of Education are difficult to meet, due to the fact that the subject is often taught by teachers not holding an informatics degree, the lack of suitable teaching material and the expectations of pupils and families, who tend to identify informatics with the use of computer applications.

In this work we study the relationship between PAC learning and the property of uniform convergence. We define the concept of polynomial uniform convergence of relative frequencies to probabilities in the distribution–dependent context. Let Xn = (0,1)n, let Pn be a probability distribution on Xn and let Fn⊂2xn be a class of events. The family {(Xn, Pn, Fn)}n≥1 is said to be polynomially uniformly convergent if, for all n, the probability that the maximum difference (over Fn) between the relative frequency and probability of an event exceed a given positive ε is at most δ (0 < δ < 1), when the sample on which the frequency is evaluated has size polynomial in n, 1/ε, 1/δ. Given at-sample(x1,…,xt), let Cn(t)(x1,…,xt) be the Vapnik-Chervonenkis dimension (VCdim) of the set (x1,…xt ∩ f | f ϵ Fn and M(n,t) the expectation E(Cn(t)/t). The results we obtain are: