António José Mendes

Pseudocode is a well‐known tool for introducing students to the basics of programming. The aim of pseudocode is to make algorithmic constructs easy to write and to understand. In this study, we propose the design of a ‘natural’ pseudocode, with keywords frequently used in the mother tongue, referring to basic computational instructions and execution flow control. The core premise underlying this study is that students will better develop their algorithmic skills if they use their own pseudocode, using their preferred keywords. This international study includes a sample of a total of 405 first‐year university students from different university degree programmes related to computer science, 60 Portuguese students, 58 Galician students and 287 students from Madrid. The same study is carried out in the corresponding mother tongue (Portuguese, Galician and Spanish) and divided into two phases. In the first, students answered a seven‐question survey on the design of the pseudocode language. The results of this phase indicate that most of the keywords generally used in traditional pseudocode are not among the students' preferences. In the second phase, a six‐question comprehension questionnaire was used to assess the students' understanding of algorithmic constructions by comparing three alternatives: the natural pseudocode in the mother tongue that they had chosen in the previous phase, the frequently used classical or standard pseudocode (also in the mother tongue) and pseudocode in English. The results show higher levels of understanding when using natural pseudocode.

Learning programming is a multifaceted process shaped by diverse factors, including effectively utilizing self-regulation of learning (SRL) skills. However, students frequently require a certain level of support to effectively engage in SRL, prompting a question regarding the optimal means of providing such assistance. The current literature is limited in identifying which types of support are most effective, and it's also noted that some students do not benefit from its usage. The underlying causes remain underexplored, and a deeper understanding of this phenomenon can offer insights into improving the design of regulatory support. This research is situated within this scope and was guided by the following research questions: RQ1) How does the use of self-regulation of learning support differ among students with high and low-performance levels? RQ2) How do students use and evaluate the self-regulation of learning support? Fifty-four students enrolled in an online introductory programming course participated in this investigation. The findings indicate that higher-performance students made more extensive use of the provided regulatory support, implying the potential utilization of this resource as a learning aid. Simultaneously, it's noted that some students either avoided or utilized this support to a lesser extent, underscoring that the mere provision of resources is insufficient for their effective usage. A qualitative analysis revealed some factors that might have influenced this behavior.

Self-regulation of learning (SRL) is an essential ability for academic success in multiple educational contexts, including programming education. However, understanding how students regulate themselves during programming learning is still limited. This exploratory research aimed to investigate the regulatory strategies externalized by 51 students enrolled in an introductory programming course. The objective was to identify the SRL strategies used by these students during multiple phases of the learning process and compare the SRL behavior of high and low-performers. The following research questions guided this investigation: (RQ1) What regulation of learning strategies are used by programming students?; and (RQ2) How do the SRL strategies used by high and low-performing students differ?. The findings demonstrate that learning to program involves complex psychological resources (e.g., cognition, metacognition, behavior, motivation, and emotion) and that students present heterogeneity in their SRL repertoire. In addition, high and low-performing students showed significant differences in how they regulate, which can contribute to understanding the factors that may contribute to learning programming. Lastly, we argue that for analyzing SRL strategies, it is necessary to consider the specificities of programming education, which motivated the development of a conceptual framework to describe the identified strategies and regulatory phases in this learning domain.

The use of computational scaffolding is a crucial strategy to foster students’ regulation of learning skills, which is associated with increased learning achievement. However, most interventions treat the regulatory processes as individual actions isolated from a social context. This view contradicts the most recent research that points to the importance of studying the regulatory phenomenon from a social-cognitive perspective, where students’ interactions influence their regulation of the learning process. This work explores these problems and presents multiple scaffolds to promote Self-regulation of Learning (SRL), co-regulation, and socially shared regulation of learning (SSRL) embedded within a computer-supported collaborative learning environment. A single-blind randomized controlled trial was performed with students (n = 71) enrolled in an online introductory programming course. Students were randomly assigned to three groups: 1) SRL-only support, 2) SRL, co-regulation, and SSRL support, and 3) a no support control group. The findings revealed that students who received regulatory support achieved higher course grades than the control group. However, only students who received SSRL and co-regulation support achieved superior performance in collaborative activities, confirming the importance of this type of regulation. Even though students did not increase in SRL aptitude, the intervention provided support for achieving higher grades in the course.

Blocks-based environments have been used to promote programming learning mostly in elementary and middle schools. In many countries, isolated initiatives have been launched to promote programming learning among children, but until now there is no evidence of widespread use of this type of environment in Brazil and Portugal. Consequently, it is common that many students reach higher education with little or no programming knowledge and skills. NoBug’s SnackBar is a game designed to help promote programming learning. This study examined students' behavior and attitudes when playing the game on their initiative. It used a sample of 33 undergraduate students enrolled in an introductory programming course. The variables studied were students' performance and engagement, satisfaction, and problem-solving strategies. The main findings were (1) better performing students had a high level of perceived learning, (2) all the students had similar perceptions about their fun while playing, (3) the leader board was the most used game element not directly related to learning and (4) the top-ranked students access previous solutions to help them solve a new mission, while the others often use a trial-and-error approach.

Social dimension plays a crucial role in the learning process. The use of technological resources to stimulate and mediate the interaction between students is known as Computer-supported collaborative learning (CSCL). Despite being widely used in programming education, the summarization of the academic literature about this topic is scarce. To create that body of knowledge, a systematic literature review was performed. The findings provide an understanding of how collaboration is explored in introductory programming, resources used to stimulate them and challenges in the process. Opportunities for future research are discussed, especially related to motivation, self-efficacy and engagement in CSCL, and the exploitation of learning analytics.