Use these original prompts to practise the different tasks in the University of Auckland MMI. Each example includes the high-yield behaviour to focus on, without prescribing one memorised answer.
Do not try to mention every possible issue. In three minutes, a smaller number of developed reasoning chains is usually more useful than a rushed list. State what matters, explain who is affected and why, then show what should happen next.
Formal terms such as autonomy, equity or cultural safety are optional. They only help when the underlying idea is applied accurately to the scenario and changes the recommendation.
These prompts are not recalled interview content. They are designed to reproduce the type of thinking each published domain can require.
A health service can send its mobile clinic to the town with the largest population or to a smaller community with much poorer transport access. What should the service consider, and how should it decide where the clinic goes?
Explain the unequal barrier and its practical effect, compare plausible allocation approaches, then connect a recommendation to participation, implementation and review.
A competent patient requests a treatment that may offer a small benefit but carries substantial risk and cost. The treating team believes a safer option is more appropriate. How should the competing considerations be balanced?
Identify the stakeholders and central tension, reach a clear or conditional position, develop the strongest case on each side and explain proportionate safeguards.
Imagine that clinical training is more emotionally demanding and administratively complex than you expected. How would you decide whether this was a temporary challenge, an area for development or a genuine mismatch with the career?
Test motivation against real career demands, use more than one source of evidence and explain what support, reflection or experience would change the decision.
A hospital is considering an AI tool to help prioritise referrals. Supporters expect faster processing, while critics are concerned about errors and unequal performance. What do you think the hospital should do?
Separate claims from evidence, compare benefits and risks, give a reasoned view and make the implementation, oversight and revision process concrete.
Describe a limitation in the way you approach difficult tasks. How did you recognise it, what effect can it have on other people and what are you doing differently now?
Use a specific but non-sensitive example, distinguish intention from impact and show an observable change rather than naming a polished weakness.
Tell us about a time you disagreed with how a group was approaching a task. What did you do, what happened and what would you carry into a healthcare team?
Make your own contribution clear, explain the information available at the time, consider another perspective and link the learning to a specific future behaviour.
A district service has lower follow-up attendance among Māori patients. Leaders propose sending more reminder messages. How should the service understand the issue and develop its response?
Avoid attributing inequity to identity, examine structural and service mechanisms, preserve patient choice and give Māori meaningful influence over design and evaluation.
This is one defensible response to the AI-supported triage prompt. It is an illustration, not a script or the only acceptable position.
I would support a limited trial rather than immediately using the tool to make final referral decisions. Patients are affected by how quickly they receive care and by the risk that a tool performs less accurately for groups that were poorly represented in its training data. Clinicians also remain responsible for safe decisions, while the hospital has an interest in reducing avoidable delay.
In favour of a trial, consistent automated screening may help staff find urgent referrals sooner and spend more time on complex cases. The strongest objection is that speed is not beneficial if errors are hidden or distributed unfairly. I would therefore keep a clinician responsible for the final priority, give staff a straightforward way to override the tool and tell patients how their information is being used.
Before wider use, the hospital should compare waiting times, missed urgent cases and false alerts with the existing process, including results for relevant patient groups rather than only an overall average. Patient and clinical representatives should help decide what level of error is acceptable. If performance is unsafe or inequitable, the trial should pause while the model or the workflow is changed. On that basis, my position is a transparent, supervised trial with clear accountability, not replacement of clinical judgement.
No. These are original practice prompts created for preparation. StationReady does not use recalled, confidential or official interview questions.
For realistic 2026 practice, use 30 seconds to prepare and up to three minutes to respond. Early in training, an untimed review can still be useful for learning how to structure the reasoning.
No. A correct concept explained in plain language should receive the same credit as formal vocabulary. Naming a principle without applying it to the facts adds little.
Check whether you answered the actual task, developed consequences rather than listing ideas, considered the strongest relevant alternative and made your recommendation practical and reviewable.
Record or type a response, then review the transparent criteria, evidence decisions and practical next step in your private report.