Lane 1: Formats for supervised assessment
To assess learning goals that must be achieved without AI, it is important to choose assessment formats that make student reasoning and performance visible and difficult to outsource to AI.
Assessment formats that verify independent mastery
A balanced assessment strategy often combines several formats to ensure that different learning outcomes are assessed appropriately. When selecting assessment formats, also consider accessibility and fairness, ensuring that students have adequate preparation and support for formats such as oral or practical assessments. Choose formats that make student reasoning and performance visible and difficult to outsource to AI.
Practical assessments: highlighting independent practice
Practical assessments are particularly robust because they evaluate performance rather than descriptions of performance, making it difficult to outsource the task to AI. Or, put differently: the lecturer can directly observe or verify the student’s independent reasoning, knowledge, or performance.
Overview of robust Lane 1 formats:
| Format | Use to assess: | Key tips |
| On-site written exam |
|
Align questions with learning outcomes and connect to course materials |
| Oral exam or presentation |
|
Use follow-up questions to probe depth of understanding |
| Practical or skills test |
|
Assess execution of tasks in real time |
| In-class activities |
|
Use short exercises or quizzes completed during class sessions |
| Validated assignments |
|
Combine with oral explanations, presentations, or exam questions to verify that the student understands and can explain the work |
Examples of Lane 1 assessments across disciplines
These are particularly robust because the examiner can probe reasoning and understanding in real time.
Examples:
- individual oral exams
- interactive oral assessments
- viva voce (defence-style questioning)
- thesis defence
- oral explanation of a method or solution
- Socratic questioning about a case study
- short oral defence of a written assignment
Why they work for Lane 1:
- students must explain their reasoning
- examiners can ask follow-up questions
- understanding becomes immediately visible.
Key differences between oral presentations, interactive oral assessments and viva voces:
| Aspect | Oral presentation | Interactive oral assessment (IOA) | Viva voce |
| Role of the assessor | Listener and evaluator; may ask clarifying or evaluative questions at the end. | Active participant in conversation who probes reasoning, asks follow-ups, and explores application. | Examiner testing depth, rigour, and defensibility of knowledge claims. |
| Level of interaction | Low to moderate; interaction is secondary and time-limited. | High and continuous; interaction is central to the assessment. | High, but examiner-controlled rather than dialogic. |
| Preparation style for students | Rehearsal of a polished, structured talk. | Preparation for conceptual understanding, reasoning, and explanation, not a script. | Deep preparation of subject matter and one’s own work, anticipating critical questioning. |
| Primary skills assessed | Oral communication, structure, clarity, synthesis of content. | Application, reasoning, adaptability, professional communication, metacognition. | Depth of understanding, academic rigour, argumentation, and justification. |
| Authenticity / real-world alignment | Variable; often academic or simulated audience contexts. | High; tasks often mirror professional conversations or disciplinary practice. | Moderate to high; aligned to academic research cultures rather than professional practice. |
| Handling of uncertainty or errors | Errors may undermine perceived quality of performance. | Errors become productive moments for probing understanding and reasoning. | Errors may be challenged directly; emphasis on academic defensibility. |
Examples:
- structured debates
- panel discussions
- policy deliberation exercises
- role-play negotiations.
Why they work for Lane 1:
- students must defend positions in real time
- reasoning becomes visible.
These assess actual performance rather than description, which AI cannot substitute.
Examples:
- laboratory experiments
- clinical or medical skills assessments
- diagnostic questioning
- engineering design demonstrations
- programming tasks completed in a controlled environment
- language proficiency interviews
- artistic or musical performances
- moot courts or negotiation simulations
- teaching demonstrations (e.g. for education students)
Why they work:
- the student must perform the task themselves.
Students complete analytical work during a supervised session.
Examples:
- analysing a dataset in class
- interpreting a graph or experiment
- solving a case study
- writing a short analytical essay during class
- responding to unseen texts or sources
- exam questions based on course-specific materials.
Why they work:
- students must demonstrate independent reasoning under time constraints.
Some assignments can still be done outside class if they include verification steps.
Examples:
- research paper + oral defence
- design project + presentation explaining decisions
- coding assignment + live code walkthrough
- policy brief + in-class debate
- written report + exam questions based on the report
Why they work:
- the verification step confirms the student understands the work.
The learning process becomes visible through multiple checkpoints.
Examples:
- research proposal → annotated bibliography → final paper
- draft submission + instructor discussion
- project milestones with presentations
- methodology explanation before data collection.
Why they work:
- the lecturer observes the development of the student’s thinking.
These are usually formative but can also contribute to assessment.
Examples:
- quizzes
- short writing exercises
- concept explanations
- peer teaching activities
- real-time problem solving.
Why they work:
- completed during class without external tools.