Click drop-down menu for workshop details.
Workshop 1- 14:00-15:30- Generative Genius or Plagiarism Peril? Rethinking Assessments in the Gen AI Era (Facilitators: Dr. Suzanne Stone and Dr. Fiona O’Riordan)
Generative AI models such as Chat GPT have rapidly advanced in their capabilities, raising concern for higher education professionals in respect of academic integrity. This workshop offers participants an opportunity to develop an understanding of how these tools work, and to reflect upon the ethical and practical considerations for designing assessment in the age of generative AI. Participants will be guided through exercises to explore how these tools can be used by learners to create assessment content based on assessment briefs. The output from these exercises will be interrogated and critiqued by the wider group. Participants can draw on insights gained on the strengths and limitations of these tools to inform future practice. This workshop is designed for faculty, instructional designers, librarians, and other academic professionals who are involved in preparing institutions and students for the rise of generative AI.
Note: No prior GenAI experience is required for this workshop.
Workshop 2- 15:30-17:00- Let’s talk about Academic Integrity (Facilitators: Michelle Tooher and Mairead Greene, University of Galway)
CCT Let's Talk about Academic Integrity Sept 2024 .pdf
We sometimes think that matters of academic integrity are black and white: a student has acted with integrity or not. Yet, when we investigate further, we find that it is not quite as simple as that. Through the use of scenarios, the facilitators will prompt discussion around the grey areas of academic integrity and how we can make it easier for students to act with integrity. We will explore academic integrity concerns such as collusion, group work, GenAI, and contract cheating. We will empower attendees to use scenarios in their own contexts to help promote discussion on academic integrity and explore the ‘grey’ with students and colleagues.
Registration for workshops is for conference delegates only.
Time* Session
09.00 - 9.30 am Registration
09.30 - 9.40 am Welcome Address from Neil Gallagher, CCT College President
09.40 - 9.50 am Minister’s Address and Launch of Conference, Minister for FHERIS – Patrick O’Donovan TD
09.55 - 10.40 am Keynote 1 from Dr. Kelly Ahuna. Life Lessons: Turning Academic Integrity Violations into Educational Opportunities
10.40 - 11.00 am Coffee/ transition
11.00 - 12.45 pm Parallel Session 1 (located in Banking Hall) & Parallel Session 2 (CCT College next door)
12.45 - 02.15 pm Lunch at Banking Hall – College Green Hotel
02.15 - 03.00 pm Keynote 2 from Dr. Zeenath Reza Khan. Fostering a campus-wide culture of integrity in the era of GenAI.
03.00 - 03.30 pm Student Panel: Navigating Contemporary Challenges in Academic Integrity
03.30 - 04.45 pm Parallel Session 3 (located in Banking Hall) & Parallel Session 4 (CCT College next door)
04.45 - 05.00 pm Conference closing
*Times may be subject to minor amendments.
Dr. Kelly Ahuna
Location: College Green Hotel, Banking Hall
Academic integrity is the cornerstone of higher education, required for upholding the value of the degrees our universities confer. Yet we know with certainty that many students cheat, and ever-evolving technologies make cheating easier and thus more prevalent. Responding to these violations has traditionally been reactive and punitive, yet it is possible to transform these violations into moments of growth.
At the University at Buffalo, we have developed an Academic Integrity Remediation program as an educational intervention after policy violations. Through guided reflection, we aim to broaden students’ perspectives about academic integrity beyond their own interests and help them understand the wider impact of academic dishonesty on others in our community. This shift in mindset helps students make different choices in the future and thereby avoid repeat integrity violations. This is good for them, good for our institutions, and good for our communities.
Chair: Greg O'Brien
Location: College Green Hotel, Banking Hall
Presentation 1: Creating a teaching activity for post-graduate students in the area of antidepressant drug development
1. Parallel Session 1 - presentation 1 - JPK Sept 2024.pdf
Professor John Kelly, Michelle Tooher, CELT, University of Galway; and Professor Shannon Harding, Fairfield University, CT, USA
In the Discipline of Pharmacology and Therapeutics, University of Galway there has been a long research-informed interest in the development of antidepressant drugs. We have incorporated this interest into a module (Advanced Neuropharmacology) for M.Sc. Neuropharmacology students (approx. 12-15) delivered towards the end of their second taught semester.
Previously, this content was delivered as a traditional 5-lecture block and assessed by an unseen essay-based question under invigilated examination conditions. On reflection, this form of assessment did not seem appropriate for the nature of the topic, and students tended to perform poorly, i.e. that there was a lack of alignment between the learning outcomes and the assessment. The COVID-19 pandemic had provided a glimpse of alternative assessment types, and coupled with growing concerns about Gen AI, we saw an opportunity to improve the student learning experience by:
· Providing students with an opportunity to work collaboratively and to improve their communication skills
· Moving away from the current stressful unseen assessment environment and develop a more rewarding assessment that would be authentic and less susceptible to Gen AI, than a simple take home essay.
In this presentation, we will outline the steps taken to redevelop this module incorporating team work (by forging a partnership with fellow students), cross-institution collaboration (by participating in a collaborative online workshop with students from Fairfield University, USA), as well as an independent research component. This incremental approach has resulted in a scaffolded, authentic assessment which includes an interactive oral situating the student in the role of pharmacology expert.
Presentation 2: Inclusive Assessment, Universal Design and Academic Integrity: The Challenges of Fairness and Academic Integrity for Disabled Students
2. Parallel session 1 AHEAD Richard Healy.pdf
Richard Healy, Research and Policy Officer; and Dara Ryder, CEO AHEAD
AHEAD is an Irish NGO that works with and for disabled people to shape inclusive and empowering environments in tertiary education and employment. This objective is underpinned by a firm commitment to the implementation of Universal Design for Learning (UDL) into national policy instruments and pedagogical practice, (AHEAD, 2024). For disabled students, the exam, as an indicator of competency has been explicated as a potential locus of disadvantage and an inequitable method of deciding who passes or fails their studies, (O’Neill, 2017). It is from this perspective that this presentation examines academic integrity, as a fair and equitable mode of demonstrating learning, while retaining academic integrity and authenticity.
The Irish student body is becoming increasingly diverse, arguably due to a variety of national funding streams and strategies that foster pathways into tertiary education for under-represented cohorts, (McCarthy et al., 2018). Moreover, AHEAD research has consistently confirmed a marked surge in the rate of participation of disabled students in Irish Higher Education (HE), (AHEAD, 2022, 2023). Normative teaching practice in HE is often buttressed by an assumption of homogeneity of all students, how they learn and how they present learning through assessment. However, academic integrity should enable all students to illustrate their learning equitably.
This presentation will examine this hypothesis from the disability perspective of assessment and how they desire to learn. By analysing a range of research projects that examined changes to assessment precipitated by COVID 19, we aim to demonstrate that alternative means of assessment can promote academic integrity from the student perspective, while still maintaining requisite academic standards.
Presentation 3: Developing Identities in Diverse Disciplines through Oral Assessment.
3. Developing Identities in Diverse Disciplines through Oral Assessment.pdf
Lucy Elvis, Lecturer; Heike Felzmann, Snr Lecturer; and Devon Goodwin, Lecturer, University of Galway
This presentation brings together three higher education lecturers from the disciplines of Education and Philosophy to reflect on their experience of developing and delivering Interactive Oral assessments as a format to capture authentic, discipline-specific skills.
These interactive orals were developed for undergraduate students at the University of Galway studying their second year of undergraduate philosophy and their final year of Children’s Studies respectively. Each instructor developed the assessment both as a response to the challenge posed by GenAI and as an inclusive alternative to empower students through an assignment that would allow them to show the skills unique to their area of study in their own voice.
All three instructors will reflect on the authenticity and inclusivity of this form of assessment, drawing on lecturers’ and student experiences: i) high levels of participation with lower rates of non-completion than other forms of assignment in the programme and module, ii) student reflections that identified this assessment form as fostering deeper learning and allowing them to express themselves more effectively than in written assessments, iii) the group oral format as a means to experience community support in the assessment situation
Challenges encountered included: i) the need for extensive preparation of students for this unfamiliar form of assessment, including the creation of a rubric ii) a small minority of students experiencing extreme anxiety with this form of assessment iii) logistical issues with the planning and delivery of the assessment format. In concluding presenters will outline reasons for continuing these assessment forms and other spaces they might be utilised.
Presentation 4: Enhancing Academic Integrity Through Meaningfulness and Building Confidence and Self-Efficacy
4. Academic Integrity_Gottlöber_3_9_24.pdf
Susi Gottlöber, Associate Dean Teaching and Learning (FAH), Maynooth University
This paper builds on a previous QQI-funded research project that investigated the impact of UDL in creating authentic assessments to enhance academic integrity with a focus on Arts students. Preliminary findings from the initial study that included a student survey, focus groups, and some individual interviews, indicated that student confidence in their academic abilities and ensuring that assessments are perceived as meaningful and relevant are critical in reducing academic misconduct. Follow up research indicated that the findings regarding self-confidence and self-efficacy align with research outside the educational domain, suggesting that increased self-efficacy and confidence diminish the likelihood of cheating.
The current project investigates these findings in more detail and focuses specifically on the relationship between academic integrity and a) the role of confidence and self-efficacy in academic integrity and strategies for enhancement through teaching and assessment practices, b) how these strategies can alleviate stress and mental health issues resulting from stress, one of the main reasons identified in the survey for committing academic misconduct, and c) the use of meaningful language when discussing academic integrity with students. The latter also addresses concerns raised about using specific language in a gatekeeping function. We will suggest that academic integrity presents an adaptive problem that is intrinsically connected to the overall personal and academic development of students and that the use of UDL, authentic assessment, and flexible deadlines can address a number of reasons identified by students for committing academic misconduct but also have their limitations.
Presentation 5: Validating Student work through Presentations and focused Q&As
5. Parallel Session 1 - presentation 5 - JKM Sept 2024.pdf
Ken Healy, Ed Tech Manager/Lecturer; James Garza; Lecturer; and Marina Iantorno, Lecturer, CCT College Dublin
Ensuring that a student’s work is their own is not a new challenge, and the growth in the power of AI has brought a new dimension to this challenge. Our approach is to require students to present their work at the end of semester, and then answer questions on the work they presented. Students are advised that there will be questions about their work and, in cases where they have prepared a separate report, are allowed to have this with them. This helps reduce anxiety and facilitates a conversation with the student about their work. Questions can be focused on key areas of the submission. This can help assure us that the student (a) did the work themselves, (b) is familiar with the research they have cited and (c) understands the concepts covered in the submission. This approach also allows students to to discuss their work in a way that would resemble a job interview. Presentations and questions are a real-world context. The presentation and question approach allows students to showcase their understanding and critical thinking skills in a practical setting. It also provides an opportunity for educators to gain insight into students' thought processes and comprehension of the material. Overall, this approach promotes a deeper level of engagement and encourages students to take ownership of their learning. It also fosters a more personalised and interactive learning experience for students and educators.
Presentation 6: Creating Monster Mayhem: Allowing Student to Explore the Strengths and Weaknesses of GenAI by Developing a Browser-based Game.
6. Monster Mayhem Presentation Sam Weiss.pdf
Sam Weiss, Lecturer, CCT College Dublin.
GenAI is still a big unknown hanging over students’ heads; what it is, how it works, and are others using it in their assignments? That last concern is compounded by both the relentless marketing surrounding the technology, as well as justifiable fear of missing out. Students may feel compelled to utilize GenAI in their assignments - not as an easy way to get ahead, but out of fear that they would otherwise fall behind.
This presentation will be a live demonstration of a case study to address the latter concern, in a way that aligns with the conference values of well-being and human-centredness. This is achieved by enhancing understanding of the technologies while addressing student concerns.
Students in a Client-Side Programming class were shown the design brief for a novel browser-based game. The initial design contains elements that are commonly found in the training sets of GenAI models. This brief was input into a GenAI (ChatGPT 3.5) and the resulting game demonstrated in front of the class. The game was largely successful.
Next, one design parameter was changed to something not well represented in the training data. This not only prevented the GenAI from completing the task, but made it markedly worse at the other aspects. Students were then instructed to finish the game, with or without the use of GenAI, and to document their process.
It is hoped that this demonstration will further this conference’s themes of Authentic Assessment, by exploring how GenAI can help/not help when coding novel solutions; and Prevention, as greater awareness of the technologies limitations may reduce the pressure on students to compromise their academic integrity.
Chair: Ruth Ní Bheoláin
Location: CCT College Dublin (beside College Green Hotel)
Presentation 1: Student and Student Teacher Perceptions of GenAI in the Classroom
Robin Flynn, Quality, Enhancement and Registration Manager; Elva Casey, Registrar; and Ronan Mulhaire Academic Integrity and Assessments Manager, Hibernia College
As microcosms of society (Battalio, 2005), classrooms act as playgrounds for emerging ideas. Engagements with GenAI in the classroom take a variety of forms from use as learning tools, assessments responses and adaptive learning routes. There is an urgent need for teachers to become acquainted with GenAI in its diverse uses by, for and with students, to determine a rationale for pedagogical and assessment decisions around its use (Chiu, 2023). It is imperative for teachers to start preparing students for a new ethical landscape (Farrelly & Baker, 2023). To do so, they must understand the attitudes of students. This paper presents the outcomes of research conducted in multiple primary schools and with student teachers, using workshops and focus groups. It explores the attitudes of both groups to engagement with GenAI and addresses three research questions;
What does appropriate use of GenAI look like for students or student teachers? What are the main immediate or future benefits and/or disadvantages of GenAI for students according to both groups? What gap exists between student and student teacher perceptions and understanding of GenAI?
The presentation will share its findings and possible implications for teacher education, in the context of a review conducted to examine what reference current teacher training programmes curricula make to academic integrity and GenAI.
Presentation 2: The Digital Academic Skills Hub (DASH) project & the impact of formative feedback on undergraduate academic confidence and capability
Katy Sutherland-Hastings, Lecturer, Birmingham City University
DASH is an interactive learning platform that provides practice opportunities and formative feedback to improve knowledge and application of core academic skills. This enhances confidence and attainment of students alongside contextualising academic integrity through digitally enhanced learning. Currently, DASH has elements focused on utilising an evidence base appropriately in academic work. This includes understanding literature and using referencing techniques correctly, with immediate feedback given.
Formative feedback opportunities enhance student experience and attainment outcomes when provided constructively to improve skills, understanding and motivation (Irons & Elkington, 2021; Winstone & Carless, 2020). Through positive impact on self-belief and confidence, formative feedback reduces academic-related stress for students (Morris, et al., 2021). A pilot was implemented on an undergraduate programme, with gathered student feedback resoundingly positive. Since the pilot, DASH is now a stand-alone digital space and designed to develop understanding regardless of previous academic experience. DASH delivers learning on a multitude of interwoven topics; referencing techniques and appraisal of journal articles are key features that align with academic integrity and application of an evidence base to academic work alongside wider contexts, such as healthcare professions.
Presentation 3: Adaptive Framework for enhancing Academic Integrity Literacy and Ethical Use of Generative AI Tools
3. Adaptive Framework for Enhancing Academic Integrity Literacy - Ml & DM.pdf
Muhammad Iqbal, Snr Lecturer; and David McQuaid, Lecturer, CCT College Dublin
In this study, we explore the current challenges and ambiguities encountered by students in discerning core knowledge and information following the advent of generative AI tools. The students used these Gen AI tools to extract the relevant material for their learning requirements and performance enhancement in assessments. It is important to educate both students and ourselves about the ethical usage and implications of Gen AI tools, as this is vital for maintaining academic integrity and standards, and for preparing learners for a successful professional life. A clear, insightful, and comprehensive framework is required to ensure the pertinency, accuracy and authenticity of information generated by AI tools. We accomplished these objectives at CCT College through a multi-faceted approach, including awareness seminars, utilizing Gen AI tools to facilitate brainstorming sessions, demonstrating diverse responses from various AI tools [1], and drawing connections to traditional methods of information retention by students, such as surface as compared to deep learning [2]. We started to educate students about the importance of literature review and analysis in developing critical thinking, alongside demonstrating the use of plagiarism detection tools (e.g., Copyleaks) and discussing the pros and cons of Gen AI tools in maintaining academic integrity. The version control system has been employed to ensure consistent contribution of students' effort to authentic assessments, thereby providing transparency for all stakeholders. A framework is developed to show the incorporation of all factors and conditions for the reliable use of Gen AI tools to enhance the literacy of academic integrity. This framework can be adapted to meet the requirements of all academic levels by tailoring the constraints to suit varying degrees of academic maturity and disciplinary expectations, and this framework will be presented as a flow chart/ block diagram in the conference presentation.
Presentation 4: Referencing? It’s not difficult et al.: creating student-informed academic writing supports
4. CCT2024_IODowd_AByrne_Parallel2.pdf
Irene O'Dowd, Digital Learning Researcher; and Ann Byrne, Digital Librarian, Hibernia College
In Hibernia College, students are expected to take primary responsibility for maintaining academic integrity in their studies. However, lecturers and support staff have an important role to play in educating students about academic integrity and helping them develop the skills needed to practise it. This presentation describes a project initiated by the Digital Learning Department (DLD) to improve the College’s online referencing supports, in response to the high volume of referencing queries being received daily by the Digital Librarian. Guided by a student-centred approach, the project consisted of a comprehensive update of the College’s core Referencing Guide and the delivery of a series of drop-in webinar workshops, where referencing queries from students could be dealt with directly and specific problem areas addressed. The presentation outlines the principles informing both the updating of the Referencing Guide and the structure of the online workshops. An analysis of library usage data and student feedback survey data demonstrates student engagement with and responses to these additional supports. The presentation also describes how the initial project has led to the development of more comprehensive and differentiated academic writing and referencing supports in collaboration with academic programmes and student support. Finally, possible future directions for the project will be discussed.
Presentation 5: Experiences Running an Undergraduate Academic Integrity Research Module
5. AcademicIntegrityModule2024-Lancaster.pdf
Thomas Lancaster, Senior Teaching Fellow in Computing, Imperial College London
This presentation will discuss my experience of developing and running the world’s first academic integrity research module aimed at undergraduate students. The Academic Integrity in STEMM module was delivered for the first time at Imperial College London in 2020, with the format, content, and assessment refined in partnership with students over three subsequent iterations. The module has attracted internationally renowned guest speakers, seen the introduction of a student research conference broadcast live online, and has led to students themselves presenting at external events. The fourth run of the module saw the incorporation of generative AI within the module, with students shown how to use tools like ChatGPT to enhance their research, whilst being mindful of the challenges of working with emerging technology. The presentation will include examples of the content developed for the module and the research produced by students working in interdisciplinary teams. It will also consider changes planned for the 2025 delivery of the module. The session is intended to inspire others looking to engage students in an academic integrity discussion that goes beyond asking them to think merely about cheating, plagiarism, referencing and misconduct.
Dr. Zeenath Reza Khan.
Location: College Green Hotel, Banking Hall
Zeenath's upcoming keynote talk, titled "Fostering a campus-wide culture of integrity in the era of GenAI," will explore the critical importance of fostering academic integrity in today’s rapidly evolving educational landscape. Drawing from her decades-long studies and experiences as a leading figure in academic integrity, Khan will introduce the IEPAR framework—Inspiration, Education, Pedagogical Consideration, Assessment Design, and Response and Restorative Practice. The talk will emphasise the need for institutions to proactively cultivate a culture of integrity by inspiring ethical behaviour, educating stakeholders on the implications of misconduct, and employing pedagogical methods that empower students to take personal responsibility. Zeenath will discuss innovative assessment designs that reflect real-world readiness and the importance of fair, transparent, and educative responses to misconduct. Amidst the challenges posed by GenAI and contract cheating, Khan will advocate for a holistic approach that integrates integrity into the fabric of campus life, ensuring that all stakeholders—from students to top management—are engaged in upholding the values that sustain academic excellence.
Moderator: Dr. Zeenath Reza Khan.
Location: College Green Hotel, Banking Hall
What is your understanding of academic integrity? And do you have any thoughts on how it can be implemented?
What, if any, are your concerns about upholding academic integrity?
What do you think is the value of academic integrity?
What works/does not in terms of promoting academic integrity, and why?
Chair: Dr. Fiona O'Riordan
Location: College Green Hotel, Banking Hall
Presentation 1: Enhancing Academic Integrity through a Whole-of-Institution Approach: The CCT Swiss Cheese Model
Ruth Ní Bheoláin, Dean of Academic Affairs; Dr Fiona O'Riordan, Dean of Teaching and Learning; and Denis Cummins, Dean of Faculty, CCT College Dublin
This paper explores CCT College Dublin's strategy for promoting academic integrity, aligning with conference themes of integrity literacy, inclusivity, and ethical technology use. We introduce the "CCT Swiss Cheese Academic Integrity Model" that is in adaptation of Rundle and colleagues’ (2020) Swiss Cheese Model of assessment design for academic integrity. The model conveys a layered approach to addressing misconduct and upholding assessment integrity, reflecting national and international best practice where ENAI and NAIN call for an institutional approach to upholding academic integrity.
Core components of the model include recognising student vulnerabilities, implementing diverse actions, fostering college-wide commitment, designing robust assessments, and educating both staff and students. Each layer, from environment and assessment design to education and quality assurance, aims to mitigate risks and foster a culture of integrity.
Initiatives include ethical Artificial Intelligence (AI) practices, aligning with conference values of ethical technology use and human-centredness. The CCT Etiquette for AI in Assessments sets clear guidelines for AI use, emphasising transparency and engagement. Continuous professional development for staff and Academic Integrity Week for students promotes ongoing awareness.
Our approach is a work in progress, emphasising that maintaining academic integrity is a whole-of-institution responsibility, extending beyond the classroom. The presentation will discuss good practices and actions in each of the layers, highlighting what we have done or plan to do. It will also address the barriers and enablers encountered or envisaged. Practical tools for supporting the detection and prevention of misconduct will be discussed.
By sharing our holistic model, we aim to contribute to the collective effort of maintaining academic standards and fostering a culture of trust and integrity in higher education.
Presentation 2: Health and Social Care Professionals Education in the Age of Generative Artificial Intelligence: Lessons from Podiatric Medicine.
2. Lessons from Podiatric Medicine FINAL.pdf
Benjamin Bullen, Lecturer in Podiatric Medicine,; and Adam McIlwaine, Lecturer in Podiatric Medicine / Plagiarism and Academic Integrity Advisor. University of Galway
Podiatry educators in the United Kingdom (UK) and Ireland are responding to the use of generative artificial intelligence (GenAI), aligned with the Academic Integrity Literacy conference theme. Conference values of authentic approaches; wellbeing and human-centredness; ethical use of digital technologies and student partnership have been expressly considered. In the UK, use of unattributed GenAI among Health and Care Professions Council (HCPC) registrants, studying at Master’s level, led to a statement in the Royal College of Podiatry’s magazine, The Podiatrist, by one presenter (BB) (1). This statement expressed, “HCPC-registered podiatrists undertaking postgraduate study should also be aware of the potential implications of academic misconduct, including abuse of AI and plagiarism... If an HCPC registrant has an academic misconduct investigation upheld against them, this must be reported to the HCPC’s fitness-to-practise department” (1, p. 28).
The Regulator for Health and Social Care Professionals in Ireland, CORU, published the Podiatrists Registration Board Code of Professional Conduct and Ethics in 2021 (2). This document dictated registrants must “always behave with integrity and honesty” (2, p. 24) and research should, “follow accepted guidelines in scientific journals concerning intellectual property, copyright and acknowledging the work of others” (2, p. 27). These statements are likely to similarly impact postgraduate podiatry students and practitioners, with implications for fellow regulated professionals.
This presentation will, therefore, highlight practical implications and consider the dissemination strategy in Ireland, imperative to educate and raise awareness of the emerging issue of academic misconduct, through the inappropriate assistance of unattributed Gen AI.
Presentation 3: Academic Integrity needs to focus on the student, as opposed to detecting cheating.
3. 20240902_CCT_NorEducation.pdf
Garnet Berry, Head of Education, NOR Education
Most students today understand that modern plagiarism checkers are so good that copying/pasting from others, self-plagiarism, and collusion is not worth the risk and can have consequences.
Meanwhile, commercial ghostwriting, where there is no reliable tool available, has continued to increase. Then, with little forewarning, Generative AI became available to everyone in late 2022.
The superpowered Generative AI ghostwriter, ChatGPT and other similar tools are shaking the cheating prevention landscape, which now includes Collusion, Plagiarism, Ghostwriting, and AI writing. Current solutions around AI Detection have been proved unreliable and therefore are not a viable solution on their own.
This presentation will review how the manual 2-step process any instructor would take to validate a student’s writing can be completed at scale to combat cheating.
A 2-Step process to Confirm the student’s understanding of what they have submitted and then Compare the submission to past work to validate it is their work.
In brief, when the student uploads the submission it is possible to generate questions based on that work. The student answers these questions in real time to confirm if they understand what they submitted..
Using researched back Authorship Analysis process it is also possible to compare the students' submission to a historical text (original piece of writing) to ensure they wrote it.
The outcome is an Actionable Insights Report that can be used as evidence as part of the academic integrity process.
Chair: Dr. Suzanne Stone
Location: CCT College Dublin (beside College Green Hotel)
Presentation 1: From Policy to Practice: Implementing University of Galway’s Academic Integrity Framework
Justin Tonra, Academic Integrity Officer, University of Galway.
In June 2022, University of Galway approved a new academic integrity policy: one that was greater in scale and scope than its previous code of practice for dealing with plagiarism. The policy considered academic misconduct in a broader and more holistic fashion, created the new role of Academic Integrity Officer, and implemented the use of Courageous Conversations, among other developments. This presentation will share and reflect on the experiences of frontline staff with responsibility for promoting and upholding academic integrity. Much of its focus will be on sharing the practical experience of implementing and operationalising the new policy and reflecting on the opportunities and challenges involved. Among the topics addressed are: staff and student training and development, academic misconduct case management, responding to the advent of GenAI, and promoting a culture of academic integrity.
The conference theme that this proposal primarily addresses is “Prevention and Detection (incl. policy frameworks).” It is aligned with many of the conference values, with greater emphasis on “Authentic approaches” and “Ethical use of digital technologies.” As University of Galway was the first Irish HEI to appoint a dedicated Academic Integrity Officer, conference delegates will be interested in hearing this presentation’s reports on practical responses to current challenges and innovations in the field of academic integrity. Moreover, the presentation will provide actionable insights and solutions for other institutions currently developing policies and strategies on academic integrity.
Presentation 2: Designing Principles of Generative Artificial Use for a Higher Education Institution
2. Silvia Benini - Designing Principles of GenAI use for a higher Education Institution.pdf
Silvia Benini, Educational Developer, Academic Integrity Unit; Mary-Claire Kennedy, Academic Integrity Lead; and Fionn McGrath, Educational Developer, University of Limerick.
Generative AI (GenAI) has presented all Universities with many challenges and opportunities in terms of developing and implementing effective good practice and governance strategies. The University of Limerick (UL) is committed to upholding the highest standards of academic integrity by all students and educators. The Academic Integrity Unit at UL have led on a project to develop a set of Gen AI principles that would be applicable across the institution. The principles (Gen AI Literacy; Integrity; Innovation; Equity; Ethical and Secure Use) were identified and developed with reference to national, European and international guidance and policy documents and were designed to reflect the roles and responsibilities of the institution, staff, researchers and students. They are designed to provide an overarching framework for all stakeholders to reflect on their use of Gen AI. The principles have been mapped to the UL Graduate Attributes to demonstrate the link of Gen AI in all aspects of teaching, learning and assessment. Developing a set of principles relating to Gen AI that are applicable to all stakeholders within a University is a challenging experience. This is due to the diverse perceptions, experiences, needs of each individual stakeholder and the fast-changing nature of Gen AI.
Presentation 3: A Feature-Based Approach for Detecting LLM Generated Texts and Verifying Authorship in Educational Settings
1. A Feature-Based Approach for Detecting LLM Generated Texts (1).pdf
Professor Patrick Bours and Thomas Nyrem Eilifsen, Norwegian University of Science and Technology (NTNU)
In this paper we investigated how contract cheating can be detected, using behavioural analysis on the writing style of a student. We focus on cheating texts written by Large Language Models (LLMs), even though the same techniques can be applied to cheating texts written by humans.
We used a 2-stage approach, where in the first stage we investigated how best to distinguish between a text written by a human versus one written by an LLM. In the second stage we extend these results to make a distinction between a text written by a specific person versus a text written by an LLM. The main difference between the two stages is that in the second stage we use a clearer view of the writing style of the specific person instead of a general “human writing style”.
We used a dataset that contained over 27.000 human written texts combined with over 46.000 LLM written texts, where 11 different LLMs were used. For the analysis we defined 41 features and used 15 different Machine Learning (ML) algorithms. For the first stage we achieved an F1-score of 0.9797, by optimizing parameters and settings. For the second stage, the performance dropped to an F1-score of 0.8761, but we noticed that for many users, the system still makes 100% correct decisions.
Our findings confirm the potential for implementing these models within educational frameworks to detect LLM-generated content and verify authorship, but some further research needs to be performed to improve performance of the models.