This research, conducted in an upper post-primary Computer Science classroom, explores how students learning Python engaged with ChatGPT. A key output of the study was the co-creation, with students, of heuristics (guiding principles) to support the ethical and effective use of ChatGPT while learning to program. These include: (i) a general set to support learning in contexts where generative AI is available, and (ii) the SAV framework (Specify the problem, Ask again, Verify the output), a specific set for engaging with ChatGPT. These heuristics are presented as working prototypes.
This poster captures our experiences of incorporating video components into assessments on Software Engineering programmes. We highlight their dual role as a tool for developing soft skills and a mechanism for verifying the attainment of learning outcomes. Our approach focuses on various types of video production tasks in varied contexts, contributing to ongoing efforts to promote soft skills development in software engineering education [1, 2].
Introductory programming courses (CS1) are widely recognized as challenging for many students. Understanding the factors that contribute to student performance in CS1 has been a long-standing focus of research in computing education. One factor often associated with course performance is motivation. However, since motivation is multifaceted, it is essential to determine which specific facets of motivation contribute most significantly to success. In this study, we validate a recently proposed instrument designed to measure social aspect, meaning, independence, and self-efficacy, within the context of a CS1 course. We compare this instrument to an established intrinsic motivation scale and students’ self-reported time use. Furthermore, we examine how these instruments and self-reported time use are related to persistence and performance in CS1. Our findings reveal that among motivational factors, perceived interest and enjoyment stand out as the most significant predictors of course performance. Additionally, perceived meaning of studies plays a noteworthy role in improving outcomes, particularly over self-efficacy. These observations underscore the value of intrinsic motivation in CS1 settings.
Code comprehension refers to the ability to read, understand, and explain code effectively. While students may successfully complete introductory programming assignments, they often do so without fully grasping the underlying code, which can lead to difficulties in more advanced classes. This work presents our approach to addressing this challenge through targeted interventions that assess students’ comprehension using personalised, AI-generated questions. We share insights from a pilot study conducted to gather initial student feedback on the effectiveness of this method.
This study presents expert feedback on an immersive virtual reality environment designed to teach 3D transformations and improve spatial visualization skills in computer graphics. Traditional 2D methods can limit spatial understanding, while the VR tool offers interactive and real-time manipulation of 3D objects. The evaluation of teaching staff and domain experts highlights the educational value, usability, and potential of this approach to enhance spatial learning in computing education.
Building community and sense of belonging in computing departments is notoriously challenging - with early-year classes that are frequently very large, students who often complete much of the work online, and significant differences in prior experience with the subject. In this paper, we describe how partnership with our main student society to create a student-led peer-support scheme, where incoming students are put in families of 8-10, with 2-3 returning students acting as parents, has helped to address this situation. We describe the challenges and impact of the scheme from both a student and department point of view, and present the research we have done with three categories of students - academic students, academic parents, and students who chose not to participate in the scheme - to explore the value students see in such a scheme, the barriers to engagement and how we can co-design plans to improve the scheme in the future. Our research suggests that many students are keen to engage in a peer-support scheme, but adequate organisational support from the school, early relationship building, and effective, broad communication are crucial for success.
This paper examines the utilisation of interactive videos in modules delivered to computing degree apprentices in a hybrid format over a three-year period. Analysing the data reveals that interactive videos have a higher rate of engagement in this context compared to non-interactive, typical videos.
This keynote aims to explore the journey of Computing Education in Ireland, beginning with its early attempts in the 1970s and culminating in the introduction of a formal Senior Cycle curriculum in 2018 (comparable to Key Stages 5 and 6) as well as some interesting attempts we tried along the way. The presentation will then turn to the current educational landscape, offering an overview of how this curriculum was designed and was implemented across schools. I will examine the challenges faced by educators as they adapt to teaching a subject that differs significantly—both technically and pedagogically—from their traditional disciplines. I will also highlight initiatives aimed at building teacher capacity and fostering self-efficacy in the delivery of Computing Education. The presentation will also consider the role that third-level institutions and chapters like the ACM SIGCSE Irish chapter have played in supporting this transition, through contributions such as teacher education programmes, curriculum development input, and collaborative professional development initiatives. We will also reflect on, and acknowledge the difficulties encountered in aligning third-level priorities and structures with the rapidly evolving needs of the school system and how EU guidelines direct the trajectory of Computing Education. Drawing upon my research and active engagement in national curriculum support efforts—including professional learning programmes and industry-linked certification, I will reflect on the practicalities of implementation, identify the most impactful supports to date, and propose ways in which third-level institutions can more effectively contribute to the continued development and success of Computing Education in schools.
Accurate modeling of student knowledge is essential for delivering timely, targeted feedback in Intelligent Tutoring Systems (ITS). Knowledge Components (KCs)—discrete units of domain knowledge—have traditionally been handcrafted by experts, a process that is both time-consuming and difficult to scale. In this work, we replicate a recent Large Language Model (LLM)-based approach to automate KC extraction called LLM-KC Identifier (LLM-KCI) and extend on it by evaluating the extracted KCs using learning curve analysis. By comparing LLM-generated KCs against expert-annotated counterparts in an introductory programming course, we demonstrate that LLMs not only match experts in capturing core concepts but also bring unique advantages: consistent identification across diverse assignments and scalability. Through static overlap metrics (Jaccard similarity, overlap coefficient) and learning curve analyses, we show that certain LLMs (e.g., GPT-4o, DeepSeekR1) produce error-rate trajectories as smooth or smoother than expert-annotated KC models, effectively isolating student learning trends. Our findings suggest that automated KC extraction can become a mainstream tool for personalized learning analytics, enabling educators to rapidly adapt curriculum and interventions at scale.
Novice students can benefit from adaptive formative assessment as it can help them overcome programming learning challenges, particularly in identifying and understanding more prevalent errors. Students are encouraged to assess themselves and grow from their errors, which in turn inspires them to pursue computer programming. This paper discusses potential areas for adaptive formative assessment in computer education, with the goal of inspiring novices in introductory programming.
The meaning of source code is captured in the syntax of a programming language. To understand a particular code fragment one must thus know the mapping of the used syntactic constructs to their meaning. Novices often lack this knowledge. To make up for this, they often resort to guessing, which leads to misunderstandings in this mapping that may be hard to unlearn. In this paper we present a didactic tool called Explicit Source Prose: source code written in what we call an Explicit Syntax. An explicit syntax makes the precise semantics it represents explicit in structured natural language. We discuss different aspects of Explicit Source Prose: it does not explain purpose, it is localizable, explicitness is a spectrum, it can be generated automatically, and it reveals typos that alter a program’s meaning. We illustrate the concept of Explicit Source Prose with examples of source code and corresponding source prose. Preliminary results of a study with ten participants indicate that novices appreciate Explicit Source Prose and that they are able to more easily understand source code when an Explicit Source Prose version is available.
As Large Language Models (LLMs) like ChatGPT and GitHub Copilot gain traction in computing education, understanding their role in introductory programming (CS1) is essential. This scoping review synthesises 38 empirical studies published between 2022 and 2024, focusing on student and educator engagement with LLMs in CS1 contexts. Following Arksey and O’Malley’s five-stage framework and PRISMA-ScR guidelines, we identify four thematic areas: (1) varied student prompting behaviours, from surface-level code copying to iterative refinement; (2) evolving educator practices, from passive allowance to guided integration; (3) assessment-related tensions, notably the “assistance dilemma”; and (4) ethical concerns around bias, integrity, and access. While LLMs support debugging and code comprehension, their value depends on pedagogical framing and learner agency. Gaps remain in longitudinal research, diverse learner representation, and alignment with curriculum frameworks. We offer practical recommendations for scaffolded GenAI integration, prompt engineering strategies, and ethical classroom use. This review supports the development of CS1 curricula that foster critical AI literacy, inclusive participation, and thoughtful engagement with human–AI collaboration in programming education.
It is well documented that automated outputs from AI systems like Generative AI (GenAI) do discriminate, can be harmful and misleading or indeed both. The concern is that those most exposed to outputs from these technologies in the classroom will be learners from Black, Global Majority and other underrepresented communities. In AI Education, the lives of young people and children will be diminished if they are continually subjected to learning material that disenfranchises them in one way or another. Therefore, the initial aim of this poster (research study) is to investigate pre-service, early career and expert teachers’ existing conceptions of fairness, accuracy, transparency and ethics (FATE) in relation to AI. The research investigates ways that AI may be incorporated into classroom practice or curriculum design equitably in computing.
As the structure and purpose of doctoral education evolve rapidly, up-to-date records of ‘how a PhD actually starts’ risk being lost. This paper provides a scholarly snapshot of those practices before they shift once again. We conducted semi-structured interviews with 22 active researchers in the Computing Education Research (CER) community who had completed their PhDs in varying disciplines, mostly computer science or computing education, in six decades (from the 1970s to the 2020s). Their PhDs were earned in twelve different countries; they now work in ten of those countries. We report on a qualitative analysis of participants’ motivations for pursuing a PhD, how they found their supervisors, and how they identified research topics. The breadth of this sample puts experiences at single institutions or countries into perspective. We found a wide range of motivations for PhD studies, including – among others – the desire to do research, a strong interest in teaching, the desire for an academic credential, and a fascination with learning. We also found a wide spectrum of practices in how supervisor-student connections are formed and thesis topics identified, ranging from the supervisor having almost total control to the student being expected to take the initiative. Collectively, our participants paint a rich picture of the beginning phase of PhD studies among people working in CER. Taken together, these findings create a baseline that future scholars can compare against as doctoral training continues to change.
Peer feedback can improve learning by encouraging deep engagement with course content, exploration of diverse perspectives, and critical thinking. For computing and software engineering students, practising peer feedback can contribute to future employability. However, incorporating peer feedback mechanisms into courses can be difficult due to problems such as reduced student motivation and trust, and the resulting low-quality feedback leads to dissatisfaction. This paper presents an extension to MarkEd, a marking and moderation tool developed at the University of Edinburgh, to facilitate high-quality peer feedback. As students write their feedback to each other, it provides them with guidance powered by a Large Language Model (LLM) on improving its actionability, clarity, and constructive tone, while discouraging them from over-relying on such guidance. Moderation by human markers is also introduced to validate peer feedback correctness and completeness. The enhanced system was trialled as part of formative assessment in a large second-year introductory software engineering course. Feedback from students and markers indicated a positive impact on learning, with most students reporting that the LLM guidance helped them reflect on their comments and learn about what constitutes high-quality feedback. Markers and a lecturer indicated a reduction in workload compared to other systems, and all groups rated the system’s usability highly, although some improvements could still be made.
This poster presents a multi-tiered thematic coding scheme for analyzing student argumentation in computer science assessments. Using a dialogic and iterative approach, we developed a structured methodology to categorize inference types across different CS contexts. Our findings highlight the importance of context-sensitive analysis and insight into how coherent students’ knowledge is.
The literature focusing on highly able programmers and in particular effective strategies to support them is limited. In the wider educational context beyond Computing, the conception of ‘gifted’ learners has been undergoing a shift with research suggesting that existing strategies have been reproducing inequalities, particularly for pupils from disadvantaged backgrounds. It has been suggested that individual interventions are not effective in supporting highly able learners. Instead, a suite of different types of interventions (academic extension; cultural enrichment; personal development and removal of financial barriers) is required. In this study we sought to develop an understanding of the strategies currently being employed to support highly able programmers in secondary schools in England. We conducted a national survey of secondary computing teachers and received 108 complete responses from across the country. Participants were asked to describe the strategies they employ to support their most able programmers. These responses were coded and assigned to one (or more) of the four types of intervention. The majority (85%) of teachers described interventions which involved academic extension, with 24% describing interventions which involved personal development. Only 1 teacher mentioned cultural enrichment activities and none discussed removal of financial barriers. The participants were also asked to rate the training they received in supporting highly able programmers. Half of the participants reported receiving no training during their pre-service teacher education and 30% reported receiving no in-service training in this area.
Technology has promised to transform teaching and learning for over 25 years, from interactive whiteboards to personalised learning platforms. Generative AI large language model (LLM) chatbots, such as ChatGPT, are the latest innovation poised to reshape computing education. However, much of the research has taken a technology-driven approach to their integration rather than a student-centred one. This scoping review examines studies published between October 2022 and October 2024, using the PICRAT framework to map student activity with LLMs as passive, interactive, or creative, and educator use as replacing, amplifying, or transforming traditional teaching methods. Analysing 32 studies across K-12 and higher education, sourced following PRISMA-ScR guidelines, we find that use of AI chatbots primarily replicates or amplifies conventional instructional patterns, with most student interactions being passive or interactive rather than creative. While these tools improve efficiency, their potential for transformative learning, particularly in K-12 settings, remains under explored. This review offers insights for computing educators seeking to integrate AI into their teaching and highlights the value of using PICRAT, not only for planning and designing AI-enhanced learning activities, but also for evaluating their pedagogical impact over time.
One of the most important aspects of students’ academic experiences is their sense of belonging, which supports their motivation, persistence, and overall wellbeing. Previous studies have shown that various complex factors, including gender, race, LGBTQIA+ membership, socioeconomic status, past experiences, and institutional culture, influence students’ sense of belonging in computing. This study compares undergraduate computing students’ sense of belonging in Ireland and Scotland, highlighting key demographic and contextual factors. We found that gender and minoritisation continue to be significant factors, but their impact differs across national settings. Our findings demonstrate that social interaction and clubs/societies membership are consistently associated with higher belonging in both Irish and Scottish contexts. This work supports the development of targeted interventions to improve students’ sense of belonging in computing, particularly across different regional and institutional environments. Future work will expand this comparison to include Wales and England to develop a broader perspective across Ireland and the UK.
This paper introduces a Gamified Feedback Loop (GFL), an in-situ feedback mechanism that provides the benefits of an automated marking mechanism. The GFL led to reduced stress levels among students through timely and relevant feedback.
