Generative AI has a wide range of impacts on how we access and use information, particularly as educational settings and perspectives differ greatly across different locations. These impacts extend to society and include impacts on intellectual and creative works and the potential infringement of authorship. Differences in institutional GenAI policies (and in funding) may create unequal access to AI tools, the potential disparity in student knowledge of AI tools, responsible uses of AI tools, ethical questions about AI tools, and uneven student knowledge of the benefits and limitations of AI tools. Generative AI introduces questions concerning academic integrity, bias, and data provenance. The training data's source, reliability, veracity, and trustworthiness may be in doubt, creating broader societal concerns about the output of the Generative AI models. This working group will conduct a landscape analysis on Global South ethical questions related to the use of Generative AI tools in higher education contexts, identifying promising principles, challenges, and ways to navigate the implementation of Generative AI in ethical and principled ways.
Automata theory is fundamental to computer science education, yet students often struggle with its abstract concepts. Digital tools have improved automata visualisation, but pedagogical approaches that support the development of problem-solving skills remain limited. This paper presents AutomaTutor 2.0, which complements our previous educational mobile application with game-based competitions to enhance student engagement and performance in automata theory learning. The competition system transforms automata construction into puzzle-solving challenges where students have to arrange automata components on a grid in order to compose an automaton that matches with a given regular language. Competitions are timed events with automated scoring based on solution correctness, completion time and attempt count. Student performance is tracked via a public leaderboard, introducing extrinsic motivation alongside the intrinsic rewards of problem-solving. We evaluated the use of competitions in teaching automata theory through a six-week experimental study with undergraduate computer science students at the University of Pretoria, South Africa. Students who participated in the voluntary competitions demonstrated higher performance improvements between assessments compared to non-participants. Survey results indicated that competition participants reported increased motivation to engage with course content and improved problem-solving skills. Our findings demonstrate that integrating game-based competitions into automata theory education can effectively complement traditional teaching methods by increasing student engagement and supporting the development of practical problem-solving skills.
Programming is a core computer science competency, and mastering its theoretical foundations and practical skills is essential for success in any computing discipline. Despite its importance, we have observed a high failure rate in the introductory programming courses at our university. Therefore, it is important to investigate the factors that influence performance in introductory programming at a university level. This study examines the relationship between factors related to motivation and learning strategies and performance in an introductory programming course within the African context. Data were collected from 138 students enrolled in an introductory programming course. The instrument used in this study is the Motivated Strategies for Learning Questionnaire (MSLQ), which was designed to assess college students' motivations and learning strategies. The study's findings reveal a statistically significant positive correlation between students' self-efficacy, intrinsic goal orientation, effort regulation, and task value, and their performance in programming. On the other hand, test anxiety showed a significant but negative correlation with performance in introductory programming. Furthermore, a stepwise multiple linear regression analysis revealed that self-efficacy, test anxiety, and extrinsic goal orientation collectively account for 24.6% of the variation in programming performance. The results suggest that instructors and students need to work on improving students' motivation and learning strategies to improve performance in introductory programming.
The CS education research community has increasingly embraced researcher positionality as an important aspect influencing our research. However, how CS educators' positionality impacts pedagogical choices is less known. In the U.S., where this study takes place, Black, Latine, Native, and Pacific Islander people are even less represented among CS faculty than among CS students in higher education. Recent work documents the benefits of culturally-relevant pedagogies in CS and strategies for implementation at the higher education level. This study investigates how CS faculty's racial/ethnic identity impacts their use of these pedagogical strategies that leverage students' culture. We interviewed 21 CS faculty at Hispanic-Serving Institutions in the U.S., five of whom identified as Latine, and report findings of a phenomenographic analysis, including benefits of Latine identity alignment, and strategies of non-Latine faculty to achieve cultural competence.
Static code-analysis metrics have long been used to assess and guide successful software projects. Traditionally, these metrics have measured software's commercial or professional suitability. This work explores six code metrics designed to provide educational insight, specifically comparing pre- and post-LLM student work. With these metrics we measure over a thousand CS1 and CS2 final-projects, completed from 2018 to 2024 inclusive. We first examine how student work has changed with the curriculum's embrace of LLMs in 2023. The results show a clear LLM-correlated impact on students' early computing work, both positive and negative. Our CS1 offers five different project themes, and we add perspective by comparing how the metrics vary across these themes. The results underscore the enduring importance of curriculum, pedagogy, and scaffolding on student work. We conclude with our program's ongoing adaptations to this mix of signals, increasing student agency and prioritizing student demonstrations and discussions of their chosen project work.
Contextualizing computer science education has been recognized as a key factor in enhancing student engagement and learning outcomes. This study investigates the initial perceptions of university computer science faculty in Africa regarding the benefits, adoption challenges, and institutional support required for the successful integration of contextually relevant materials into introductory computer science (CS1) courses. Faculty then assessed a set of previously developed contextually tailored materials, grounded in Banks' Additive Approach to curriculum reform and aligned to the CS curricula 2023. The research adopted qualitative methods, gathering data through open-ended surveys from 22 CS faculty across 9 African countries. Thematic analysis identified key patterns in the responses from faculty, who generally expressed positive perceptions of integrating contextualized materials. They agreed such materials could enhance engagement without distracting from core objectives, but emphasized the need for careful integration. Insights from faculty highlighted that successful implementation requires substantial institutional support, including curriculum reform, textbook development, and faculty training, with universities playing a critical role in adoption.
Algorithms are a fundamental part of computer science education, with expressing and tracing forming key aspects of their study. Both pseudocode and code are used to express algorithms. Pseudocode helps abstract away specifics of the programming language when specifying an algorithm, but it can introduce ambiguities due to its informal nature and it seldom supports a structured way to trace an algorithm. On the other hand, tracing a program provides a precise description but can conflate algorithmic logic with language-specific details, obscuring the inherent structure of an algorithm. In this experience report, we propose employing the Mapcode framework for expressing and tracing algorithms. Mapcode represents an algorithm as a set of transformations between well-defined spaces, naturally producing execution traces that are independent of programming constructs. Unlike traditional notional machines, which model the execution of programs, Mapcode machines model the high-level execution of algorithms. Mapcode traces reflect underlying logic and are independent of programming-language-specific control structures like while loops. We report preliminary findings from two classroom activities which explored the application of Mapcode to express and trace algorithms. In the first activity, third-year undergraduate students drew Mapcode Execution Diagrams to trace simple algorithms already specified in the Mapcode framework. We found that a majority of students were successful in constructing the diagrams. In the second activity, students in a Principles of Programming Languages course successfully defined algorithms from various paradigms (greedy, dynamic programming, sorting, etc.) as Mapcode machines and implemented them in Python and Racket.
This paper explores the current and potential impact of Artificial Intelligence (AI) on higher education from the perspective of computer science faculty and students across three universities in Accra, Ghana. With participation from 248 students and 21 faculty, the study employed surveys and focus group discussions to capture participants' perceptions on AI, how AI is currently integrated into teaching, learning and assessment, and the impact of AI on these activities. The results show that both students and faculty perceive that AI is transforming higher education, and faculty are exploring AI integration in their work. However, many faculty have concerns about the negative effects of students' use of generative AI and continue to educate them on responsible use of AI for learning. Such concerns influence how faculty assess student learning reverting to in-person assessments among other strategies. Overall, the consensus is that AI is transforming higher education, and there is a need for institutions, faculty and students to adapt to this reality.
The delivery of culturally relevant content in K-12 computing education is widely recognized as essential for engaging students and fostering meaningful learning experiences. However, in many African contexts, K-12 computing teachers feel unprepared to address this need. Course materials often fail to reflect the cultural and linguistic diversity of the students they serve. In this study, we conducted participatory design sessions with education stakeholders, including primary and secondary school teachers, to develop culturally relevant programming activities for a national programming outreach initiative in Botswana. Our findings emphasize the importance of relatable examples and a balanced approach to integrating cultural elements into learning materials. Challenges, such as diverse dialects and difficulties with direct translation, led stakeholders to adopt a bilingual approach to the initiative. This research contributes insights into the complexities of integrating culturally relevant programming languages and course materials and highlights the need for flexibility and adaptability. The implications of this study include the need for cross-regional research collaborations, teacher training and guidelines, and empowering teachers to effectively integrate cultural elements, promoting inclusivity and cultural responsiveness in programming education.
Integrating artificial intelligence (AI) tools into software engineering (SE) has transformed industry workflows, presenting new opportunities and challenges for computing education. This experience report details the systematic redesign of two advanced undergraduate courses-Capstone (CS30) and Software Engineering (CS31)-to incorporate AI tools into software design, implementation, and testing phases while preserving core engineering rigor. The redesign was structured to ensure that AI tools enhanced rather than replaced students' problem-solving and SE skills, reinforcing critical engagement with AI-generated outputs. Unlike introductory courses, where AI may overshadow fundamental learning, this intervention introduced AI at a stage where students had already developed programming and software engineering competencies. The curriculum emphasized project-based learning, iterative refinement, debugging strategies, and structured AI tool comparisons, encouraging students to critically assess AI-assisted workflows against traditional manual development through assignments and reflections. While AI tools improved efficiency in repetitive tasks, challenges such as hallucinations, biases, and prompt engineering needs highlighted the importance of scaffolding AI-assisted learning. This report presents lessons learned, key challenges, and best practices for integrating AI into computing curricula, ensuring students balance AI automation with rigorous problem-solving and preparation for industry workflows.
With artificial intelligence (AI) topics entering computer science school curricula, there is a growing need for instructional approaches that support the teaching of data-related concepts and practices. In academic AI and data science programs, an established teaching and learning method is the data case study. However, despite its widespread adoption, reports on its didactic development in school settings are lacking. This article presents findings from a research project that theoretically and empirically identified the challenges of implementing the data case study in schools. Based on these findings, the method was further developed for secondary education. The result is a specification of the data case study as an action-oriented, learner-centered approach aligned with the goals of AI education in schools and designed to promote data-based judgment and problem-solving skills.
A programme integrating course (PIC) helps students acquire an overall picture of the education and thereby a better understanding of the importance of each individual course, supports students in making informed choices both during the education and thereafter, and gives them the opportunity to influence the development of the programme. PICs also help teachers gain insight into how students perceive their studies. Additionally, they help the programme director obtain student evaluations of the programme with a 100 per cent response rate. The foundation of the course consists of quarterly seminars in small cross-year groups, led by a mentor who is one of the teachers in the programme. In each seminar, students reflect both in writing and orally on a given theme relevant to their education and their studies, such as study techniques, procrastination, professional roles, minorities and equal treatment, internationalisation, educational quality, ergonomics, and mental health. We report our experiences from programme integrating courses in two computer science and engineering programmes at KTH Royal Institute of Technology for 15 years and 3 years, respectively: which challenges in the programmes motivated the introduction of PICs, challenges in starting and running PICs, and students' and teachers' experiences. Most students appreciate the programme integrating course, and their appreciation increases over the run of the three-year course.
Recent reports emphasize the need for a shift in undergraduate Computer Science (CS) education towards competency-based learning and the integration of ethical and responsible computing practices. Community-based service learning (CBSL) is a proven strategy to achieve these goals. However, despite a strong track record, service learning can be challenging to implement and may not always guarantee successful student experiences. We introduce Scaffolded Projects for the Social Good (SPSG), a framework resulting from 10+ years of iterative refinement of running externally sponsored team-based projects in a software studio environment. SPSG aims to guide educators through the complexities of CBSL by addressing critical factors that have been consistently cited as barriers to the successful adoption of CBSL: project scoping, skill matching, managing project timelines that extend beyond a single term, community partner relationships, and project handoff and maintenance. This paper outlines SPSG's benefits over traditional methods and provides a sampling of student experiences in several recent SPSG projects. We reflect on how the framework can enhance student competencies through iterative formative feedback associated with a comprehensive set of project deliverables.
