ABSTRACT Background and Context Despite the growing demand for computing professionals, high dropout rates and pronounced gender disparities persist in CS programs. While factors such as grit and programming experience influence student retention and success and may differ by gender, how these differences shape retention remains unclear. Objective This paper employs the Social Cognitive Career Theory (SCCT) to study how gender differences in grit and programming experience across the introductory computing curriculum affect students’ intention to complete their computer science (CS) degree. Method We collected survey data from 303 undergraduate CS major students enrolled in CS1, CS2, and CS3 at an engineering-focused university in the southeastern United States. Findings Results showed differences between men and women at different course levels in terms of grit and programming experience, which in turn influenced their self-efficacy to study CS, outcome expectations, attitude toward computing, and consequently their intention to continue studying CS as their major. Implications The study highlights the need for intervention strategies to enhance retention and provide tailored, gender-inclusive support for computer science students.

ABSTRACT Background and Context As programming becomes increasingly popular for children, non-formal learning environments such as code clubs play a vital role in fostering their interest and engagement in computing. However, ensuring equitable access and sustained participation remains a significant challenge. Objective To identify best practices in educational and management strategies employed by European code club practitioners to advance access, equity, and inclusion in non-formal computing education. Method A mixed sampling approach was applied to conduct semi-structured interviews with 17 code club practitioners from eight European countries. The data analysis process followed Bingham’s five-phase qualitative analysis, drawing on deductive and inductive coding strategies. Findings Our findings highlight diverse approaches practitioners employ to reduce participation barriers, support learners’ engagement and emotions, and create inclusive learning environments. Despite these efforts, challenges persist in reaching and retaining underrepresented groups, underscoring the need for strong approaches to equity-centered designs. Implications The equity-related strategies and interventions identified in this study contribute valuable insights and inspiration for policymakers, educators, and practitioners in both non-formal and formal educational settings seeking to promote social justice in their pedagogical approaches, organizational structures, and policy development.

ABSTRACT Background and Context Data structures and algorithms (DSA) play a central role in programming. However, misconceptions related to DSA, i.e. students’ mental models which differ from the expert thinking, have been studied to a moderate extent. Concept inventories (CI) are validated tests which measure students’ knowledge utilizing misconceptions. CIs enable teachers to tailor their teaching methods and identify areas where students struggle. Objective This systematic literature review investigates what CIs exist on DSA topics and what misconceptions these CIs contain. Method We performed backward and forward snowballing to collect 52 papers discussing DSA & recursion CIs. The scope was finally reduced to nine reports which (i) describe DSA misconceptions and (ii) have conducted at least one pilot test in the CI development process. Misconceptions were extracted from these reports using their explicit definitions or by interpreting CI questions. Findings We identified seven DSA CIs in the final set of nine reports. We have extracted 92 misconceptions or students’ difficulties from these reports and mapped them onto ACM/IEEE-CS/AAAI Computer Science Curricula 2023. Implications Our catalog of 92 DSA misconceptions and difficulties helps instructors gain pedagogical content knowledge and improve learning materials. Future research should investigate misconceptions on DSA subtopics not covered here but present in the curricula recommendation. Reviews and CI development could be extended to the 22 reports of misconceptions finally excluded from this study.

ABSTRACT Background and Context Computational thinking (CT) is increasingly recognized as a critical skill in today’s technology-driven society, necessitating its integration into K-12 education. While numerous teacher professional development (PD) programs exist, gaps remain in our understanding of the effects of online asynchronous PD followed with classroom implementation on teacher self-efficacy. Objective This study examines how an online, curriculum-based PD followed with classroom teaching, can enhance elementary teachers’ self-efficacy for integrating CT, offering broader insights for design of effective CT teacher education programs. Methods A total of 48 elementary teachers participated in a randomized controlled trial study. Teachers either completed PD on CT and game programming – implementing these activities in their math classrooms – or taught math as usual. Pre- and post-data were collected on teachers’ self-efficacy for teaching and integrating CT, and confidence in coding and teaching elementary math. Qualitative data elucidated factors contributing to and constraining teachers’ self-efficacy and the support they needed. Findings Findings showed that, after two semesters of PD and teaching 2 weeks of game programming, teachers had a significant increase in self-efficacy for teaching and integrating CT, as well as an increase in coding confidence. Four key factors influenced their confidence: (1) Gaining experience teaching and familiarity with content, (2) Learning new pedagogical approaches to support students’ progress, (3) Witnessing positive student interest, understanding and engagement and (4) Working with curricular materials, and receiving technical and logistical support. Implications These findings offer insights on how to structure elementary CT teacher education and PD programs to enhance elementary school teachers’ ability to integrate CT effectively.

ABSTRACT Background and Context Traditional text and blockbased programming often limits the development of computational thinking (CT) skills, while educational robotics (ER) provides a handson alternative that can better foster these skills. However, how robotics environments promote CT remains unclear. This study focuses on Chinese high school students engaged in programming activities using ER. Objective This study aims to develop and validate an integrative model that explains learner-centered and technology-centered determinants of students’ CT skills. Method A cross‐sectional empirical study was conducted with 339 high school students who completed both a questionnaire and a programming fundamentals test. We applied Partial Least Squares Structural Equation Modeling (PLSSEM) to assess the measurement and structural models, Importance-Performance Matrix Analysis (IPMA) to identify priority factors, and Artificial Neural Network (ANN) analysis to capture non‐linear relationships and rank variable importance. Findings Task‐Technology fit (TTF), learning motivation (LM), cognitive engagement, and prior programming knowledge each exhibited significant positive effects on CT, with LM (β = 0.333, p < .001) and TTF (β = 0.251, p < .01) emerging as the most influential factors. The integrated model accounted for 54.1% of the variance in CT. IPMA and ANN results corroborated the centrality of motivation and fit in enhancing CT. Implications By extending the Cognitive Affective Theory of Multimedia Learning through incorporation of TAM and TTF constructs, this study highlights the dual importance of aligning robotics features with task requirements and fostering student motivation.

ABSTRACT Background and Context Many media have been used to teach computing in a culturally responsive manner, including art, dance, philosophy, and e-textiles. Objective Computational embroidery (CE), which involves programming an embroidery machine to embroider designs onto fabric, is a promising new medium, but prior work has yet to explore how it shapes learning and instruction. Method To explore how this new medium-shaped learning and teaching, we taught a six-week CE class to high school students that explored culture, embroidery, and computing. We used qualitative thematic analysis of student and instructor reflections and student work. Findings We found that students leveraged CE to make the work they wanted to while entangling their identities and their cultures. CE enables students to learn computing by doing meaningful work, which supports their personal programming growth and identities. Implications These findings demonstrate the potential learning opportunities of using CE to facilitate meaningful, culturally responsive learning experiences that promote technical, creative, and cultural literacy.

ABSTRACT Background and Context Generative artificial intelligence (AI) tools such as ChatGPT are emerging as influential resources in education, with potential to support learning in high school programming through interactive, on-demand assistance. Objective This cross-sectional study explored relationships between high school students’ ability to complete programming activities when permitted to use generative AI, their interest and attitude towards programming, and foundational mathematical knowledge. Method High school students undertaking a computing course at an independent boys school in Sydney, Australia were invited to complete an online survey and attend a lab-based session to complete seven programming-related tasks of varying complexity. Findings Eighty-six students consented; analyses of programming performance were restricted to 64 (74.4%) students who attended the lab-based session, with mathematics grades available for 63 students. Among lab participants, mean assessment score was 29.86 ± 18.91 (out of 100); interest and attitude towards programming score was 56.69 ± 20.08 (out of 92); and math score was 79.95 ± 16.88 (out of 100). There were no statistically significant differences in the assessment scores across groups with high (n = 30), moderate (n = 29), and low (n = 5) interest and attitude towards programming, or across groups with math grade A (n = 23), B (n = 27), and C or below (n = 13). Implications Generative AI can help high school students complete programming activities, even among those without high interest and attitude towards programming or strong mathematical skills. These findings suggest that, when carefully integrated into teaching, generative AI tools can broaden student engagement in programming, while cautioning against overreliance on AI-generated solutions.

ABSTRACT Background and context Enabling students to make informed decisions about the use and design of computing technology is an internationally proclaimed goal of K-12 computing education (CE). Designing related activities and assessments requires a better understanding of the structure of competent sociotechnical decision-making. Objective The paper aims to develop a competency structure model for sociotechnical decision-making in K-12 CE. Method Based on an inductive content analysis of existing competency models from K-12 science education, the resulting codes were used in a secondary analysis of 35 interview transcripts from a previous study on the conceptual structure of computing impacts. Findings The analysis resulted in a structure model comprising 11 (sub-)competencies, their hierarchical and lateral relations, and associated knowledge, skills and attitudes. Implications The developed competency structure model can inform the design and targeted scaffolding of sociotechnical decision-making activities and the formulation of related action-based learning objectives.

ABSTRACT Background and context Some skills taught in introductory programming courses are categorized into 1) explaining code, 2) arranging lines of code in correct sequence, 3) tracing through the execution of a program and 4) writing code from scratch. Objective Knowing if a programming skill is a prerequisite to another would benefit teachers in properly planning the course and structuring the order in which they present activities relating to new content. Prior attempts to establish a skill hierarchy have suffered from methodological issues. Method In this study, we used the conviction measure from association rule mining to perform pair-wise comparisons of five skills: Write, Trace, Reverse trace, Sequence and Explain code. We used the data from four exams with more than 600 participants where students solved programming assignments of different skills for several programming topics. Findings Our findings matched the previous finding that tracing is a prerequisite for students to learn to write code. Contradicting the previous claims, our analysis showed that using the mean threshold writing code is a prerequisite to explaining code. However, there is no clear relationship when we change the threshold to the median. Unlike prior work, we did not find a clear prerequisite relationship between sequencing code and writing or explaining code. Implications Our research can help instructors by systematically arranging the skills students exercise when encountering a new topic. The goal is to help instructors properly teach and assess programming in a fashion most effective for learning by leveraging the relationship between skills.