Assessment, Ako and Assurance: Lessons from the Waikato Forum

Assessment in higher education is coming back to design. The questions getting attention now are the old ones: what a task is actually for, and where the student is when they sit down to do it.

More than seventy delegates from nineteen institutions came to The Pā in Hamilton for the University of Waikato x Cadmus Learning & Assessment Forum, a full day on the future of learning and assessment in Aotearoa. Seven of New Zealand's eight universities were represented, alongside polytechnics and student associations. 

Underneath the day's questions sits ako, the Māori concept that teaching and learning form a reciprocal relationship with students as imparters of knowledge, not just recipients of instruction  (further reading). It surfaced session after session, raised by students, practitioners and leadership alike.

Five threads ran through the day: what students told the room, the case for reducing before redesigning, the shift from paper level to programme level, trust as a design problem, and what changes when the process of learning becomes visible.

Students Opened the Day, and Set the Agenda

The forum was framed by a panel of student representatives discussing their experiences with assessment and generative AI. 

• How has student life changed since most people in the room completed their studies?

• Where is the line between using AI in assessments and cheating?

• What are the realities of assessment workload, delivery and security for students?

Workload came first. At one of the universities represented, most students work 20 to 30 hours on top of study, and deadlines set independently across four papers arrive together. A paper, in New Zealand usage, is a single course, and nobody is looking across the four.

On integrity, students were candid on where AI replaces their own critical thinking and reasoning. What is missing, they said, is being taught to use these tools well, in the way the library once taught research method.

One student described a lecturer who prohibited AI in his paper, then presented AI-generated slides containing a fabricated case. The point made of it was that trust runs in both directions, and that the lecturer and his students had reached for the same tool for the same reason, which is that they ran out of time.

Later sessions reached the same conclusion from the institutional side. Categories built to describe AI use had quietly become a way of policing the amount of it.

Blanket prohibition of AI drew the most scrutiny in disciplines whose professions have adopted these tools at scale. Law came up by name, where AI is barred outright in some papers that are preparing students for an industry investing heavily in it. What the student panel asked for was explanation - tell a class why a task exists and what thinking it is meant to build, and the boundary starts to make sense from the inside.

Distance students carry a sharper version of the same problem. Securing a remote assessment usually means a verification process, and a verification process asks for resources not every student has and a view into places some would rather keep private. Those requirements can be entirely legitimate and still cost some students far more than others.

What students said they wanted more of was authentic assessment, designed to teach graduate-ready skills. Those were the tasks they named as the most enjoyable, and where more learning occurs.

Over-assessment

Assessment volume came up in every session, raised independently by students, practitioners and university leaders.

Two waves of over-assessment have built up since the pandemic. The first was designed to draw students back into classrooms, and the second went in as a control on generative AI. Both were rational approaches at the time, but without programme-level coordination have resulted in unsustainable assessment workloads for students and increased instances of students disengaging with learning.

Examples of papers with daily assessments, papers with more than twenty distinct assessment tasks, and exam numbers at one university close to doubling in a few years illustrated the burden that students are under. 

The response taking shape is reduction ahead of redesign. One university has a policy in governance that would bring assessment down to three items per paper, phased over two years and starting at 100 level, alongside a university-wide conversation about learning outcomes and constructive alignment.

The institutional panel closed by turning the standard back on the room. A candid remark highlighted how academics being assessed every six years by the Performance-Based Research Fund often bemoan the experience. In stark contrast, students are assessed at something closer to weekly, and that is treated as routine.

From Paper Level to Programme Level

The two-lane model is fast becoming the settled direction across the New Zealand sector. In short, it splits a programme into secure assessment at identified verification points, and authentic, accessible assessment everywhere else. 

A programme view is what makes the model work. Looking across a whole qualification means rethinking the purpose and placement of every assessment in it. In some cases, a paper may legitimately hold no assessment at all, because assurance of learning happens at identified points elsewhere. External accrediting bodies set the outer limits of where those points can sit, so every programme map comes out looking different.

The institutions retiring their intermediate AI frameworks describe the move to two lanes as a progression along a path they were already on, with the framework having done its job of giving staff and students a shared vocabulary in the meantime. Leading the change was described as a tricky dance and at times a clumsy one, and one institution said plainly that it has no magic bullet.

Trust as the Foundation of Design

If the structural work sets where assessment sits, Professor Kathryn Sutherland’s session asked what makes it effective, and it turned the question the sector has been asking since 2022 around.

Institutional attention has gone almost entirely into whether students can be trusted. The research presented asked instead what evidence there is that students trust us, and which teaching practices generate it. Trust in a classroom was defined as a student's willingness to take risks on the judgement that their teacher is committed to their success, and the research is clear that it is not evenly distributed. It varies by social class, by ethnicity, and by whether a student is the first in their family to be there.

The framework that came out of the work has four dimensions. Command of the field, which is where most students start, because the institution made the appointment. Demonstrated care for the individual student. Integrity and fairness enacted in the marking instead of asserted in a policy. And acknowledgement that the room is not made up of people like the teacher, with design that follows from that.

What it asks for is small and immediately available. Learning students' names, providing practice opportunities, co-designing rubrics and explaining the purpose of a task instead of leaving students to infer it. 

The session also named an asymmetry that later speakers returned to unprompted. Disclosure of AI use is asked of students continuously and volunteered by academics rarely. Underneath it sat the reciprocal relationship between teaching and learning, and that the two are entwined closely enough that designing assessment has to be a conversation with the people being assessed. Ako again, from the other side of the desk.

Making Learning Visible

Practitioner sessions explored assessment redesigns that optimise for learning and assurance over detection through:

1. Prioritising process over final product: When the stages of the work are assessed alongside the finished piece, the evidence of learning comes from how the student got there.
2. Using process analytics to assure learning and support students: Data about how a piece of work developed provides educators with opportunities to support students pre-submission, and assurance that students have met learning outcomes. 
3. Scaffolding: Breaking a large task into smaller supported steps lets students build and practise each capability the final piece will ask for, before it counts.
4. Incorporating oral assessment tasks into written pieces: Asking students to talk through their work alongside submitting it surfaces understanding that a written piece on its own cannot show.

By shifting the focus of where educator time and effort is expended in assessment tasks, students are given the opportunity to better engage in the learning process, rather than optimising for a final product that can be easily created with AI. 

The Question Underneath the Redesign

Beneath the sessions on policy and practice, a recurring question arose in many discussions. 

What does generative AI mean for knowledge itself, and what follows for an institution whose currency has been the production of written work. 

Put in its hardest form, what does a student get from a university that they could not get from an online provider or an AI teacher. Although the question remains open, it is a reflective moment for institutions.

Assessment built on relational pedagogy, ako and good design, with verification of learning at a programme level, offers institutions the opportunity to redirect its focus from security and policing back towards learning and student success - arguably one of the core tenets of tertiary study. 

A Collective Effort

No institution is solving this alone. What is emerging is a sector-wide shift from simply securing assessment to redesigning it, built on programme-level thinking, visible learning processes and a willingness to challenge current practices. Timelines everyone agrees are tight, scaling questions around oral assessment and uneven capability across teaching teams are versions of the same problem, and the institutions in the room were able to discuss these challenges candidly. 

At Cadmus, we are proud to support this work. Our platform makes the process of learning visible through assessment design, analytics, oral assessment and group work in one place, and we support institutions worldwide to enact assessment reform, implement policy and uplift assessment practices. 

Get in touch with Cadmus 

Watch the Day Back

Read the full report · Watch the sessions

Ngā Mihi

Our sincere thanks to the University of Waikato, Te Whare Wānanga o Waikato, for welcoming us into The Pā and for the pōwhiri. Thank you to the Waikato team who made the day possible.

Thank You to Our Speakers

We would also like to warmly thank everyone who presented. The forum was a reminder of how much progress happens by putting the right people in one room for a day.

• Professor Kathryn Sutherland, Centre for Academic Development, Te Herenga Waka Victoria University of Wellington
• Professor Jennifer Milam, Deputy Vice-Chancellor Academic, University of Waikato
• Associate Professor Kirsten Petrie, Assistant Vice-Chancellor Academic Excellence, University of Waikato
• Stephanie Reid, Associate Director, Learning and Teaching Design, Waipapa Taumata Rau University of Auckland
• Jean Jacoby, Director of Education Futures, Te Kunenga ki Pūrehuroa Massey University
• Professor Mick Grimley, Dean Future Learning and Development, Te Whare Wānanga o Waitaha University of Canterbury
• Josh Burrell, Head of Product and Capability, Toi Ohomai Institute of Technology
• Professor Jason Stephens, Academic Head, School of Learning, Development and Professional Practice, Waipapa Taumata Rau University of Auckland
• Hye Ji (Erica) Lee, Associate Director, Learning and Teaching, Te Ara Whakawhiti Massey University College
• Regan O'Malley, Associate Director, Learning and Teaching, Te Ara Whakawhiti Massey University College
• Megan Clune, Professional Teaching Fellow, Waipapa Taumata Rau University of Auckland
• Bronwyn Alton, Principal Academic Staff Member, Toi Ohomai Institute of Technology
• Clare Crawshaw, Principal Academic Staff Member, Toi Ohomai Institute of Technology
• Herk Kailis, Founder and Co-Chief Executive Officer, Cadmus

And a special thanks to the students who opened the day. Your insights and candor were invaluable in framing the forum’s discussions. 

• Meighan Hay, Advocacy Lead, Waikato Students' Union
• Seamus Lohrey, President, Waikato Students' Union
• Matthew Lee, Undergraduate Education Vice-President, Auckland University Students' Association
• Ayla Brook, General Representative Distance and Pāmamao, Te Tira Ahu Pae, Massey Students’ Association

Learn more about Cadmus