Purpose of this guide
This guide supports markers working with open (Lane 2) assessments which are designed as assessment for learning and where any AI use is allowed and tasks are completed in unsupervised environments. It is intended as a framing document instead of a checklist.
We recommend using this guide to focus marking and feedback on learning, judgement, and development, rather than inferring provenance or detecting AI. Apply rubric criteria in ways that make disciplinary standards visible, and provide feedback that students can use to improve before later secure (Lane 1) assessments. Remember, it’s the role of secure assessments (not open assessments) to verify students’ capability, so open assessments liberate you and your students to focus on growth.
When marking open assessments, your role is not to:
- determine whether a student “really” did the work themselves,
- infer learning solely from writing style, fluency, or polish, particularly where AI tools may have shaped these features of the work,
- provide final proof of achievement of course or program learning outcomes.
Your role is to:
- make disciplinary standards visible,
- look beyond surface features of the submitted work to assess the quality of reasoning and judgement demonstrated,
- offer informed judgement that helps students calibrate and develop their own understanding of quality, and
- provide feedback students can use in their learning approaches and assessments including secure (Lane 1) assessments.
Marking secure assessments is, in many ways, easier because it often involves spotting correct or incorrect responses. In contrast, open assessments prioritise learning and judgement development.
It is understandable that marking open assessments may at times feel like “marking the AI”. This guide is intended to support markers to focus on the quality of reasoning and judgement demonstrated, rather than its provenance.
The broader rationale for this approach to open assessment and feedback is discussed in the article “Why feedback matters (and matters more) for open assessments.”
AI tools are often very good at producing work that has high levels of polish. Therefore, in developing open assessments, unit coordinators may instead wish to focus on criteria, aligned to learning outcomes, such as:
- Conceptual connection to the task – how ideas are selected, related, and integrated in response to the specific question or problem being addressed
- Reasoning and justification – how claims or decisions are explained
- Quality of judgement – how choices are made, prioritised, or defended
- Effective use of AI tools – if relevant, how AI is used as a partner to shape and improve the work including how its responses are constrained, evaluated, or rejected
While Section 2 outlines what markers should focus on, this section explains how rubric design can support those judgements and feedback.
Rubrics for open assessment should be designed to help provide actionable feedback rather than just to assist in grading. As one aim of these assessments is to increase students’ ability to judge the quality of the work themselves, rubrics should be constructed with this in mind and be made available to students. Ideally, time would also be spent in class going through how they can self assess their work before submission and understand the feedback afterwards.
Effective criteria include:
- describing what students should do with ideas,
- requiring judgement or decision‑making, and
- critically engaging with AI as a thinking partner.
Conceptual integration
How effectively are disciplinary ideas used to make sense of the task?
- Not meeting (Fail range): Concepts are absent, incorrect, or used in ways that do not address the task.
- Developing (Pass range): Concepts are identified but treated largely in isolation.
- Proficient (Credit–Distinction range): Concepts are connected to explain relationships or patterns.
- Strong (High distinction range): Concepts are integrated to generate insight and shape understanding.
Reasoning and justification
How well are claims or decisions explained and supported?
- Not meeting (Fail range): Claims are asserted without explanation, justification, or relevance to the task, such that the intended learning outcomes related to reasoning are not demonstrated.
- Developing (Pass range): Claims are stated with limited explanation.
- Proficient (Credit–Distinction range): Claims are supported with relevant reasons or evidence.
- Strong (High distinction range): Judgements are explicit, well justified, and weighed against alternatives.
Contextual application and transfer
How effectively are ideas adapted to the specific task?
- Not meeting (Fail range): Ideas are not applied, or are applied in ways that do not align with the task context or intended learning outcomes.
- Developing (Pass range): Ideas are applied generically.
- Proficient (Credit–Distinction range): Ideas are adapted appropriately to the context.
- Strong (High distinction range): Ideas are reshaped thoughtfully, with assumptions and implications made explicit.
Evaluative judgement
How well does the student judge quality (of their own or others’ work)?
- Not meeting (Fail range): There is insufficient evidence of evaluative judgement aligned to the criteria or standards relevant to the task.
- Developing (Pass range): Evaluation is minimal or descriptive.
- Proficient (Credit–Distinction range): Strengths and limitations are identified using criteria.
- Strong (High distinction range): Quality is weighed against standards, with priorities for improvement identified.
Effective use of tools (including AI) (where relevant)
- Not meeting (Fail range): Tool use undermines the task goals or is contrary to the intended learning outcomes.
- Developing (Pass range): Tool use is uncritical.
- Proficient (Credit–Distinction range): Tools are used purposefully to support disciplinary thinking.
- Strong (High distinction range): Tool use is strategic and critically evaluated against task goals.
The examples below show how commonly used assessment criteria can be elaborated, rather than replaced, to make expectations more explicit and to support students’ ability to judge the quality of their own work.
| Traditional criteria | Elaborated criteria |
| Written work | |
| Demonstrates understanding of key concepts | Uses disciplinary concepts to explain relationships, implications, or tensions relevant to the task. |
| Quality of argument | Makes and justifies judgements, showing how claims are prioritised and supported. |
| Critical analysis | Evaluates ideas or perspectives using explicit criteria, including strengths, limitations, or alternatives. |
| Academic writing quality | Clarity supports reasoning; coherence of thinking |
| Case studies | |
| Identification of key issues | Identifies and prioritises issues based on relevance, impact, or constraints. |
| Application of theory | Adapts theoretical ideas to the specific case, explaining how context shapes decisions. |
| Problem solving | Makes and justifies decisions in response to the case, considering alternatives or trade‑offs. |
| Recommendations | Evaluates recommendations in light of evidence, assumptions, and consequences. |
| Data analysis | |
| Correct analysis techniques | Selects and applies analytical approaches appropriate to the data and purpose. |
| Accuracy of results | Interprets results to inform judgement; notes limitations or uncertainty. |
| Presentation of data | Uses representations purposefully to support interpretation and decision making. |
| Discussion of findings | Explains what the analysis means for the problem or question, not just what was found. |
| Experimental design | |
| Appropriate methodology | Justifies design choices in relation to the research question and constraints. |
| Control of variables | Evaluates trade‑offs and limitations in design decisions. |
| Feasibility | Judges practicality and robustness of the design under realistic conditions. |
| Innovation | Makes purposeful design choices rather than reproducing standard approaches. |
| Research analysis | |
| Breadth of sources | Selects sources strategically and explains their relevance to the research focus. |
| Literature review quality | Synthesises sources to frame judgement, not just demonstrate coverage. |
| Critical engagement | Evaluates contributions, gaps, or tensions in the literature. |
| Referencing accuracy | Sources and results inform judgement. |
| Portfolio or journal | |
| Completion of entries | Demonstrates purposeful selection and development of work over time. |
| Reflection on learning | Evaluates decisions, strategies, or outcomes using explicit criteria. |
| Evidence of progress | Identifies how feedback or experience informed change or refinement. |
| Personal insight | Articulates transferable insights about practice or judgement. |
| Presentation | |
| Clarity of communication | Communicates ideas clearly to support reasoning and judgement. |
| Structure and organisation | Structures content to illustrate priorities and key decisions. |
| Engagement | Adapts content and emphasis to audience and purpose. |
| Use of visual aids | Uses visuals purposefully to support interpretation, not decoration. |
| Debate / conversation | |
| Participation | Contributes purposefully to advance understanding or discussion. |
| Knowledge of topic | Uses disciplinary knowledge to respond, challenge, or build on others’ contributions. |
| Argumentation | Makes and defends positions in response to context and counter‑arguments. |
| Listening skills | Adapts responses based on engagement with others’ ideas. |
| Evaluation | |
| Identification of criteria | Selects and justifies criteria appropriate to the evaluation context. |
| Assessment of effectiveness | Weighs strengths and limitations against chosen criteria. |
| Conclusions | Makes defensible evaluative judgements, acknowledging uncertainty or trade‑offs. |
| Recommendations | Aligns recommendations clearly with evaluative reasoning. |
| AI supported work | |
| Originality | Shapes, evaluates, and adapts AI outputs using disciplinary judgement. |
| Independent work | Demonstrates independent judgement in selecting or rejecting AI tools and material. |
| Digital skills | Uses AI purposefully to support thinking. |
| Ethical practice | Weighs ethical and professional implications of AI use. |
Open assessments are where students learn how to judge quality themselves, regardless of the source of the material. Supporting evaluative judgement also requires attention to students’ feedback literacy – their capacity to interpret feedback, manage emotional responses, and use comments to improve future work. The feedback should consider how it is:
- providing comments which focus on next steps and what to try next time,
- helping students with likely emotional responses to the feedback by communicating care,
- naming what distinguishes adequate from strong work,
- explaining standards explicitly
- reinforcing that judgement is learned, not assumed.
Open assessments do not carry the burden of assurance or verification of students’ capability. That responsibility sits with secure assessments across the program. That’s a good thing because it means the marking is about learning and feedback rather than passing or failing a student. Marking open assessments matters because it helps students understand standards, calibrate and develop their own judgement, and improve before repeating a similar task or before attempting secure assessments.
Further reading
The article Why feedback matters (and matters more) for open assessments explores the rationale for open assessment, feedback for learning, and evaluative judgement in AI‑rich environments.
