Assessment Symposium 2026

---

Keynote: Scenario-Based English Language Assessment and Human–AI Partnerships for Supporting International ESL Students in the Pre- and Post-Entry University Context

Abstract
How can assessment improve the quality of the learning experience for students? And how might rethinking assessment open up alternative and more effective ways of teaching from the student perspective? These guiding questions frame this keynote’s exploration of English language assessment in higher education. As universities increasingly rely on standardised English language test scores for admission, growing evidence suggests that such scores are insufficient predictors of international students’ academic language performance beyond entry.

This keynote argues for a reconceptualisation of pre- and post-entry English language assessment through scenario-based, integrated-skills tasks that more closely reflect the communicative, disciplinary, and cognitive demands of university study. Drawing on research in authentic assessment, integrated language assessment, and post-entry language support, the presentation proposes a model in which AI-mediated and teacher-mediated feedback operate in complementary ways.

The keynote also foregrounds the urgent need to support international students’ AI literacy, enabling them to use generative tools critically, ethically, and strategically rather than dependently. While AI demonstrates strengths in pattern recognition and surface-level language support, it remains limited in evaluating creativity, originality, voice, and nuanced academic judgment, which are areas where English language teachers play an irreplaceable role. Ultimately, this keynote argues that equitable pre- and postentry language development depends not on replacing teachers with AI, but on designing assessment systems that integrate technological affordances with pedagogical expertise, fostering student agency, academic success, and linguistic confidence in multilingual higher education contexts.

Bio
Aek Phakiti is a Professor of TESOL at the University of Sydney. His research expertise lies in language testing and assessment, with particular emphasis on test development and validation, strategic competence, test-taking processes, and the cognitive and metacognitive dimensions of language performance. In recent years, his work has extended to integrating artificial intelligence into language assessment, including automated evaluation and feedback systems that support both formative and summative practices.

Professor Phakiti has an extensive publication record in the field, including Assessment for Language Teaching (Cambridge University Press) and Language Testing and Assessment: From Theory to Practice (Bloomsbury Academic). His research continues to shape contemporary approaches to English language assessment, particularly in the areas of authentic, integrated-skills assessment and the alignment of assessment with learning in higher education contexts.

Find out more:

LinkedIn: https://www.linkedin.com/in/aek-phakiti

---

Keynote: Rethinking authentication: Supporting diverse students for high-stakes oral assessment in the age of AI 

Abstract
Generative AI has destabilised written assessment and driven a rapid turn toward interactive oral assessments, valued for being difficult to outsource. The University of Sydney now requires Honours (and soon PhD) candidates to pass an interactive oral demonstrating authorship of their own work; for coursework, interactive orals have also been widely adopted as a form of secure assessment. This turn should concern every English teacher preparing students for university, because the interactive oral does two jobs at once that are easily confused. As an authentication instrument it asks whether a student did the work; as authentic assessment it measures a genuine graduate capability: communicating and defending ideas under pressure in English. The first construct treats spoken fluency as irrelevant variance to be controlled; the second treats it as the point. Left unexamined, this conflation authenticates integrity while quietly disadvantaging EAL, international and first-generation students, who carry added real-time cognitive load for reasons that may have nothing to do with whether they did the work.

This presentation argues that the AI-driven turn risks collapsing the interactive oral into a gatekeeping instrument, which is an apt framing for authentication, but a limiting one for coursework, where the oral is better understood as a learnable professional genre our field is well placed to teach. Drawing on our observations of Honours orals, collaboration with Faculty, and the support workshops we delivered, we identify where students genuinely struggle and how scaffolding shifts oral assessment anxiety from corrosive to productive. We then advance an AI-symmetry argument: the technology that broke written assessment is also what makes oral assessment teachable at scale. We demonstrate an AI practice agent we developed: a private, judgement-free rehearsal partner that lets students fail and retry a high-stakes spoken genre as often as they need.

We close with a principle: If we require students to pass an oral to authenticate their learning, we owe them the means to learn that genre.

Bios
Dr Adam Gall is a Learning Success Advisor in the Learning Hub, Academic Language and Learning team at the University of Sydney. His work supports students to develop their academic skills, including writing and study strategies, through individual consultations, embedded and cross-disciplinary workshops, and digital learning resources. He has been leading the design of student support for the oral assessment requirements in Honours and PhD cohorts.

Dr Alexandra García is a Senior Lecturer (education-focused) leading the Learning Hub, Academic Language and Learning at the University of Sydney, a central unit supporting students in academic literacies and research skills across all levels and Faculties. With a strong background in linguistics and education, her work spans academic practice and research, focusing on the reach and impact of academic interventions, including those linked to assessment. She has been selected as a Horizon Education Fellow for a project investigating and designing student support for assessment in the age of AI. She is currently collaborating on designing student support for the oral assessment requirement in Honours and PhD cohorts.

LinkedIn – https://www.linkedin.com/in/alexandra-garc

With thanks to our Event Sponsors

---

Registration and Tea/Coffee
Location: Foyer Area Level 4 (D18.04.C400)
9:30 – 10:00 AEST

Acknowledgement of/Welcome to Country
Welcome to the UECA Assessment Symposium
Location: Lecture Theatre 308 (D18.03.308)
10:00 – 10:15 AEST

Associate Professor Simon Winetroube, UECA President and Director, Curtin English

Benchmarking Project Presentation
Location: Lecture Theatre 308 (D18.03.308)
11:15 – 11:45 AEST

Lunch
Location: Foyer Area & Upper Garden Level 4 (D18.04.C400 C400B)
11:45 – 12:25 AEST

Session 1A – Reading with AI: Keeping it Critical
Location: Lecture Theatre 308 (D18.03.308)
12:25 – 13:00 AEST

Lisa Mangion & Kerry Moisidis, UOW College

Abstract
The increasing availability of generative AI presents both opportunities and challenges for assessment design in English for Academic Purposes (EAP) programs. This paper reflects on the design, implementation, and iterative revision of an “Article Comparison Task” in a university pathway program for international students.

In this assessment, students were required to choose a topic and select and compare two related academic journal articles, explicitly using AI tools to support reading comprehension, vocabulary development, and idea generation. While the task aimed to promote critical engagement with academic texts and transparent, ethical AI use, outcomes were mixed. Some students demonstrated enhanced conceptual understanding and synthesis skills, while others relied heavily on AI outputs, raising questions about authorship, depth of engagement, and the validity of written products as evidence of learning. In response, the task was revised across multiple iterations.

Building on this reflective cycle, the paper proposes a shift towards an interactive, seminar-style speaking assessment. In this new format, students respond to guided questions about their chosen articles in a live tutorial setting, enabling assessors to probe understanding, evaluate critical thinking in real time, and triangulate evidence of learning beyond AI-mediated writing. This presentation contributes practical insights into rethinking assessment in AI-rich contexts. It argues that integrating spoken interaction offers a promising pathway for enhancing assessment validity and invites discussion on these possibilities.

Bios
Lisa Mangion is Subject Coordinator of the English language pathway program at UOW College. She has over 10 years’ experience teaching ELICOS. As one of the Deputy Convenors of the English Australia Assessment SIG, she actively contributes to sector-wide discussions on assessment best practice.

Kerry Moisidis is the Assistant Program Manager at UOW College Australia with over 30 years of ELICOS teaching experience. As a former Critical Literacy subject coordinator, she developed practical approaches to strengthen international students’ critical thinking skills and support their academic transition.

Session 1B – Rethinking Assessment Support in the Age of AI
Location: Seminar Room 304 (D18.03.304)
12:25 – 13:00 AEST

Emma Morton, Hawthorne Melbourne and Monash University

Abstract
Group assessment remains a common but contested feature of English language education. While it is often used to encourage collaboration, communication, and peer learning, students frequently report dissatisfaction with group tasks because of uneven contribution, unclear accountability, perceived unfairness, and limited control over final outcomes. These issues also create challenges for teachers, who must design fair tasks, monitor group processes, respond to student concerns, and provide meaningful feedback within increasingly heavy assessment workloads.

At the same time, universities are increasingly promoting AI-empowered and external academic support tools as ways to provide students with additional feedback, writing guidance, and study support. Such tools may also help teachers manage workload by supporting formative feedback, identifying common areas of difficulty, and helping students prepare more effectively before submission. However, generic support tools cannot resolve assessment-design problems unless they are aligned with the specific tensions embedded in the task.

This presentation uses group assessment as a concrete example to examine the relationship between assessment design, student experience, teacher workload, and AI-empowered support. Drawing on student feedback and reflections on assessment practice, it argues that rethinking assessment requires more than adding digital feedback around existing tasks. Instead, assessment support should be designed around the actual sources of difficulty: fairness, contribution, accountability, learner agency, and teacher capacity. The presentation offers practical implications for using AI-empowered support to strengthen assessment design.

Bio
Emma Morton is a computational linguist, English language educator, and researcher in AI-mediated language learning. She has taught and held academic roles with the University of Cambridge, the University of Melbourne, and Navitas. Her research examines task-oriented and open-ended AI dialogue systems, lexical development, feedback engagement, learner agency, and responsible AI use in advanced English language learning. Her work connects computational linguistics, applied linguistics, writing pedagogy, and language education.

Session 1C – Assessments Design Students Can’t Outsource to AI
Location: Seminar Room 322 (D18.03.322)
12:25 – 13:00 AEST

Vlad Hyvaa Terentyev, Impact English College

Abstract
The rapid advancement of generative AI poses significant challenges to traditional assessment practices in higher education, particularly those emphasising information recall, formulaic structures, and reproducible outputs. Such formats are increasingly vulnerable to outsourcing, undermining academic integrity and obscuring evidence of genuine learning. This presentation responds by proposing a shift from reactive AI detection strategies to proactive AI resilient assessment design, which prioritises authentic engagement, critical thinking, and visible learning processes.

Drawing on practical experience in EAP, IELTS preparation, and university level teaching within multilingual academic contexts, the session introduces a suite of classroom tested strategies for designing assessments that minimise the incentive and feasibility of AI outsourcing. These include: staged assessment tasks with interim feedback; reflective components (e.g., learning journals, writer’s memos); oral follow ups (viva voce); collaborative problem solving; in class drafting sessions; personalised, context specific prompts; and structured feedback cycles requiring revision statements.

Each strategy is illustrated with concrete examples and shown to achieve three key outcomes: (1) increased student engagement through interactive and relevant tasks; (2) strengthened academic integrity by valuing process over product; and (3) enhanced development of transferable skills (critical thinking, communication, collaboration, metacognition) required beyond the classroom.

Rather than framing AI as a threat, this approach reimagines it as a catalyst for pedagogical innovation. Attendees will leave with immediately actionable ideas for designing assessments that foster authentic participation and deeper learning, applicable across EAP and higher education contexts.

Bio
Vlad Hyvaa is a DELTA-qualified English language educator based in Melbourne with extensive experience in university-level EAP, IELTS preparation, academic literacy, and interdisciplinary teaching. As a Fulbright Scholar at University of Nebraska–Lincoln, he further developed his interests in intercultural communication, learner autonomy, and academic engagement. He also spent eight years teaching a wide range of university subjects at Ningbo University, working with multilingual students across diverse academic contexts. His professional interests include assessment design, speaking and writing development, student transition into higher education, communicative language teaching, and creating inclusive, student-centred learning environments that support both academic achievement and real-world communication.

Session 2A – Validity & Value in EAP Workshop – PART 1
Location: Lecture Theatre 308 (D18.03.308)
13:05 – 13:40 AEST

Cara Dinneen, Macquarie University College
Dr Agnes Bodis, Macquarie University

Abstract
The rapid proliferation of gen.ai has done something useful: it has made the validity problems in our EAP assessments impossible to ignore. If a student can produce a competent literature review, annotated bibliography, research essay, or memorised presentation with minimal independent language production, what are these tasks actually measuring? And where they ever really measuring what we thought?

This workshop invites EAP assessment designers to interrogate two interconnected questions. First, a question of validity. In a gen.ai environment, are our assessments still measuring the constructs that matter? Participants will share their current assessment tasks and adaptations they have already made in response to gen.ai.
Second, a question of value. Is the time investment in these tasks, for students and for teachers, proportionate and sustainable? As the field of EAP moves toward process-oriented assessment with multiple feedback touchpoints, we need to ask whether our assessment designs are working smarter or simply harder. Participants will sketch out the marking demands of their current assessment tasks and collaboratively explore design principles that preserve validity without compounding workload.

By the end of the session, participants will have collaboratively established principles on what we need to keep doing, start doing and stop doing in EAP assessment.

Bios
Cara Dinneen is Education Manager for English Language and English Medium Instruction Programs at Macquarie University College and Convenor for the English Australia Assessment SIG. She holds a Master of TESOL, Graduate Certificates in Educational Research, and Business Educational Leadership, and has 25 years’ experience in teaching, learning and leadership.

Dr Agnes Bodis PhD, SFHEA, is a Senior Lecturer and Course Director of the Applied Linguistics and TESOL courses at Macquarie University. She has over 25 years’ experience in English language teaching, teacher training and curriculum and assessment design. Her research focuses on linguistic inclusivity in education, language policy, and language assessment.

Session 2B – Providing a toolkit for academic success
Location: Seminar Room 304 (D18.03.304)
13:05 – 13:40 AEST

Ellie Young, University of Sydney Centre for English Teaching

Abstract
The Centre for English Teaching (CET) embarked on the creation of a new Direct Entry Course curriculum in 2023 that would meet the evolving needs of international students about to commence their studies at the University of Sydney. This evidence-based curriculum was comprehensively planned and developed over two years, and began to be delivered at the end of 2025.

One unique aspect of the CET curriculum is the ‘Academic Success Toolkit’ (AST), which consists of a set of six essential academic skills: research skills, critical thinking, digital literacy and AI, self-regulation and monitoring, feedback literacy and groupwork skills. One key concern for the new curriculum was ensuring that these skills were meaningfully woven into lessons and assessments to ensure students were confident and capable of meeting emerging challenges when studying in their university degree programs.

This presentation will explore the AST and how it has been enacted in speaking assessment tasks, as well as in speaking lessons that are constructively aligned to the formative and summative assessment tasks. We will also unpack teacher and student perspectives on how the AST has been adopted and enacted. From the vantage point of a curriculum writer, teacher, and reviewer of teacher feedback, we can see how theory and practice have come together and evaluate how this new approach sets students up for future academic success.

Bio
Ellie Young is a teacher and curriculum writer at the Centre for English Teaching. With 21 years of English language teaching experience, she is passionate about designing and facilitating engaging learning experiences that support student success and help learners develop the academic skills needed for university study.

Session 2C – From memorised scripts to authentic voices: Our journey rethinking reflective assessment in the AI era
Location: Seminar Room 322 (D18.03.322)
13:05 – 13:40 AEST

Zdenka Anderson & Suraya Rahme, Macquarie University College

Abstract
Many university programs require students to engage in reflective thinking. Our MQ College Direct Entry English program therefore aims to develop this skill before students begin their degree studies. At the same time, as a language program, we must ensure that assessment reflects students’ own language ability, which has become increasingly complex in the age of generative AI. This presentation shares our journey of redesigning a reflective assessment task, drawing on both a senior teacher’s curriculum perspective and a classroom teacher’s experience implementing the task.

The assessment was first introduced in 2020 as a closed-book reflective writing task to ensure authenticity. However, many responses became formulaic or memorised, limiting deeper reflection. In a second iteration, the task was redesigned as a multimodal reflective journal. As GenAI use among students grew, however, this format made it difficult to determine students’ actual language ability and the extent of AI involvement.

The task has now been redesigned as a two-part assessment: written reflections (non-assessed, where AI use is assumed), followed by an individual viva voce discussion to verify authorship and assess students’ ability to explain and extend their ideas. This session shares practical insights into maintaining reflective assessment, supporting the development of students’ reflective skills, and keeping a clear focus on language development in the AI era.

Bios
Zdenka Anderson holds a Master’s degree in Education (TESOL) and currently works as a Senior Teacher in the English Language Programs (ELP) and English-Medium Instruction (EMI) programs at Macquarie University College. She has over 25 years of experience teaching General and Academic English and Direct Entry pathway programs in higher education. Her professional interests include curriculum development and assessment design. She also contributes to the design and delivery of professional development and training for educators teaching in EMI programs.

Suraya Rahme holds a Master’s degree in TESOL from The University of New South Wales and has over 20 years of experience teaching university students in Australia and overseas. She has taught oral communication, sophomore rhetoric, academic English, argumentation, direct entry programs, and workplace culture courses and has mentored instructors in TESOL programs. She currently teaches at Macquarie University and The University of Sydney.

Session 3A – Validity & Value in EAP Workshop – PART 2
Location: Lecture Theatre 308 (D18.03.308)
13:45 – 14:20 AEST

Cara Dinneen, Macquarie University College
Dr Agnes Bodis, Macquarie University

This is Part 2 of the workshop, continued from Session 2A (Part 1).

Session 3B – Student Success Through Feedback Literacy
Location: Seminar Room 304 (D18.03.304)
13:45 – 14:20 AEST

Suria McGee, English Language Centre, Adelaide University

Abstract
Teachers invest significant time and effort in providing detailed feedback, yet gaps often remain between feedback received and student improvement. This presentation explores an inquiry into student feedback literacy within an English language pathway context, focusing on students’ ability to interpret, evaluate and apply feedback to future learning.

The inquiry emerged through teacher reflection and collaborative investigation undertaken through action research, which served as the process for examining feedback practices and student engagement with feedback. Teachers reflected on common challenges, gathered evidence from classroom practice, explored different forms of feedback, and incorporated student voice to better understand how students experienced and used feedback.

Findings highlighted significant challenges related to student feedback literacy. The inquiry revealed important gaps between teacher intentions and student understanding of feedback processes. In response, practical changes were made to feedback design and delivery, including colour-coded rubric-linked feedback, focused micro-marking approaches, and explicit feedback strategies to support student understanding. Targeted teacher development sessions were also introduced to support greater consistency in feedback practices across teaching teams.

Participants will leave with practical examples of small, implementable changes to feedback practice, alongside strategies for supporting student feedback literacy that can be adapted and trialled within their own educational contexts.

Bio
Suria McGee is a Senior Learning and Teaching Advisor at Adelaide University’s English Language Centre and has contributed to various programs including general and academic English, pathway programs, teacher training, and English as a Medium of Instruction (EMI). She has taught in both Australian and international contexts, with a strong focus on education quality and inclusive education. Suria is passionate about fostering diversity, inclusion, and student wellbeing while supporting initiatives that enhance the learning experience of international students.

Session 3C – Building towards writing assessment success
Location: Seminar Room 322 (D18.03.322)
13:45 – 14:20 AEST

Simon Cosgriff, Curtin University

Abstract
The assessment of writing in English language direct entry programs requires learners to develop control over language, structure, cohesion, and genre-specific conventions. While these are addressed in typical programs, many students remain underprepared to demonstrate these skills in summative tasks. This gap often reflects limited, systematic formative preparation, where feedback is fragmented or too general to support the development of these core writing skills. At the same time, increasing pressures on teacher workload can limit opportunities for detailed, high-quality feedback, reducing its effectiveness in supporting skill development. This session presents a structured, staged approach linking a sequence of formative in-class and out-of-class tasks to the development of key writing skills assessed in summative tasks.

The model adopts a dual-feedback approach, combining independent drafting, supported by AI as a tool for formative feedback, with guided feedback cycles and targeted instruction. While AI-supported feedback can offer immediate, personalised support and encourage learner independence (Crompton et al., 2023), it is most effective within a broader pedagogical framework. Teacher feedback remains essential for developing the higher-level writing skills required for direct entry programs, particularly genre conventions, organisation, and cohesion (Eibinger et al., 2025). The approach emphasises purposeful feedback cycles that support revision, clarify expectations, and build control over key features assessed in summative tasks, enabling students to develop autonomy while benefiting from targeted feedback that supports achievement (Hübner et al., 2026). This combination helps students engage more actively in the writing process and build confidence and flexibility in demonstrating these skills in assessment.

Bio
Simon Cosgriff is a Senior Teacher at Curtin English. He has nearly 30 years’ experience in English language teaching and has worked as a teacher, teacher trainer, academic coordinator, and assessor across Japan, South Korea, New Zealand, and Australia. He holds a Doctor of Education from the University of Southern Queensland and a Master of Applied Linguistics from Macquarie University. His interests include teacher development, reflective practice, and assessment. His doctoral research explored the use of digital technologies to support autonomous professional development among native English-speaking teachers in South Korea.

Afternoon Tea
Location: Foyer Area & Upper Garden Level 4 (D18.04.C400 C400B)
14:25 – 15:00 AEST

Closing
Location: Lecture Theatre 308 (D18.03.308)
15:45 – 16:15 AEST

– First-time Presenter Award & Interstate Presenter Award
– Feedback Survey

Registrations open on Monday, August 29

Register Here

The closing date for registrations is COB 09 July, 2026.