This research has three prongs, with each comparing open- and closed-book exam questions across six years (2017-2023) in a final year undergraduate applied machine learning course. First, the authors evaluated the performance of numerous LLMs, compared to student performance, and comparing open and closed book exams. Second, at a micro level, the examination questions and categories for which LLMs were most and least effective were compared. This level of analysis is rarely if ever, discussed in the literature. The research finally investigates LLM detection techniques, specifically their efficacy in identifying replies created wholly by an LLM. It considers both raw LLM outputs and LLM outputs that have been tampered with by students, with an emphasis on academic integrity. This study is a staff-student research collaboration, featuring contributions from eight academic professionals and six students.

In 2021 CSINC (a TU Dublin research group) launched their online Computer Science MOOC called CSLINC(Computer Science Inclusive Learning Environment). CSLINC provides a suite of eight week courses in computer science (CS) which are scaffolded for teachers to support their delivery in the classroom Quille22. The uptake for CSLINC was significant with 230 teachers delivering material to 12,000 students. The goals of CSLINC are to attempt to change CS perceptions and increase the number of students choosing Leaving Certificate Computer Science (LCCS) and also to increase teacher programming self-efficacy in the delivery of CS materials, thus leading to more teachers considering to deliver the formal curricula. Previous research conducted in the area of teacher programming self-efficacy found that programming self-efficacy is lower for formal-curricula second-level CS teachers compared to CS1 Students Faherty21. In the study teachers were attending CPD aimed at those teaching or planning to teach the LCCS subject at upper second level. The purpose of this pilot study is to do an initial comparison between the programming self-efficacy of teachers who have used CSLINC (who do not deliver formal CS curricula) in the classroom compared to the teachers who are or were planning to deliver LCCS (formal curricula).

In 2018 Ireland launched a new, nationwide Computer Science curriculum for upper second-level students. The biggest challenge preventing uptake of the subject is a low supply of teachers qualified to teach the subject resulting in most schools not offering the subject, preventing most students from having the choice in the first place. Currently, approximately 150 secondary schools (of 723 nationally) offer the subject. Most school teachers who aspire to teach CS do not have a degree in computing and typically have very low self-efficacy in programming, as previously reported in the literature. There is a need to help these teachers to confidently de-liver the national CS curriculum at second-level in Ireland. CSLINC is a platform and MOOC offering teachers a scaffolded approach towards building their confidence for formal CS content delivery. CSLINC provides courses that teachers can deliver to junior second-level students before the official, national CS curriculum is offered at senior-level. This poster provides overview of the benefits of this system for second-level teachers in Ireland, which will also be of value to the computing education community internationally.

Teacher self-esteem has been found to impact student learning in a number of non-computing fields. As computing slowly becomes a part of the upper secondary school (high school) curriculum in many countries, instruments designed to measure teachers' programming self-esteem can help inform classroom practice and processes such as teacher professional development needs. This study examines if there are differences in programming self-esteem (using the Bergin Programming Self-Esteem Instrument) between upper secondary school teachers and CS1 students in Ireland. In addition this study provides evidence of validity when using this instrument(originally developed for CS1 students) to measure upper secondary school teacher programming self-esteem. To test for evidence of validity, we compared the results of the programming self-esteem construct given to upper secondary school teachers (n=130) to a recent study of programming self-esteem among CS1 students (n=693). We found evidence of both reliability and validity with teachers that aligns with the evidence found for the CS1 students, demonstrating utility for use with teacher cohorts. Comparing these findings, teachers reported statistically significantly lower programming self-esteem compared to CS1 students. Interestingly CS1 students identifying as male had a statistically significant higher programming self-esteem than those identifying as female. However, we found no statistically significant difference for teacher gender, unlike previous work. Our results indicate that teacher programming self-esteem should be given consideration in the design and implementation of professional development.

This poster describes the use of online technology to deliver K-12 teacher professional development (PD) during the COVID-19 pandemic in Ireland. Traditionally these sessions are delivered in person, with a focus on hand-on activities, but the sudden changes faced by the closures in Ireland required an alternative approach for delivering these sessions. The PD session presented in this poster was a more technically challenging micro:bit workshop, which was delivered online using the micro:bit classroom. This is typically used as an in-class, one to many instructor tool, and trialing this as a PD collaborative tool, was a novel approach. This poster presents the delivery and methodology of the session, the collaborative online format, and feedback from the participants.

This paper outlines the longitudinal development of a K-12 outreach model, to promote Computer Science in Ireland. Over a three-year period, it has been piloted to just under 9700 K-12 students from almost every county in Ireland. The model consists of a two-hour camp that introduces students to a range of Computer Science topics: addressing computing perceptions, introduction to coding and exploration of computational thinking. The model incorporates on-site school delivery and is available at no cost to any interested school across Ireland. The pilot study so far collected over 3400 surveys (pre- and post- outreach delivery). Schools from all over Ireland self-selected to participate, including male only, female only and mixed schools. The no-cost nature of the model meant schools deemed "disadvantaged", to private fee-paying schools participated. Initial findings are very positive, including the balance of male and female participants, where in the 2017-18 academic year it was 56:44 and in 2019-20 (to date), it is 35:65 respectively. Once the model is validated and tweaked (based on survey data), the model will be published (open access) for other institutions to implement the model locally. In addition, the authors intend to link schools (that the team have worked with over the three years) with local institutions, thus developing a sustainable ecosystem for the program to continue. This paper describes the model structure and outlines early findings.