Practising Interviews Alone: Drills That Still Work

Most interview preparation happens alone. No study group, no medical family, no friend who has been through it, no mock circuit within three hours of where you live. If that is you, the honest position is this: solo practice has a ceiling, and almost nobody gets anywhere near it before the ceiling matters.

What you cannot build alone is the unpredictability of a real person. What you can build alone is nearly everything else: structure, timing, fluency, the habit of taking a position, and the ability to start talking about something you have never seen before without freezing. Those are drillable, and most candidates never drill them.

The two ways solo practice goes wrong

The first is silence. Reading a prompt, thinking through a good answer and nodding along to yourself is not practice. It skips the only difficult part, which is converting a thought into ordered speech in real time. If your practice can be done on a bus without moving your lips, it is not practice.

The second is comfort. Unlimited thinking time, a prompt you chose, a second attempt when the first one wanders. All of that trains you for conditions that will not exist. A drill has to have a constraint or it is just daydreaming with extra steps.

The core drills

The cold prompt

Write a large batch of prompts on separate slips, put them in a container, and draw one at random. Read it once, take the reading time your format allows, then talk for the full station length without stopping. No pausing the timer, no restarts. Our overview of what an MMI is in Australia and New Zealand describes the usual timings, though these vary, so check the university's current admissions page for your interview and set your timer to match.

The randomness is the whole point. Choosing your own prompt lets you avoid the categories you are bad at, and those are precisely the ones that will show up.

The playback audit

Record every station. Then watch it back twice: once on mute to see your body, and once with your eyes closed to hear your voice. Splitting the channels stops you from being generally embarrassed and forces you to notice specific things.

Count, do not vibe. Count filler words. Count how many seconds passed before you said anything with a position in it. Count how many times you said the word obviously. Numbers change behaviour in a way that a vague sense of it was a bit rambly never does.

The thirty second plan

Draw a prompt and give yourself thirty seconds to say only your plan out loud: the three things you will cover and the order. Then stop and draw another one. Ten of these takes five minutes and trains the single most useful reflex in the format, which is organising before speaking.

Candidates who ramble almost never have a content problem. They have a sequencing problem, and this is the drill that fixes it.

One story, five questions

Take one experience from your own life and answer five different questions with it: teamwork, failure, communication, leadership, something that changed your mind. Say each version out loud, and force yourself to emphasise a different part each time.

This is how you stop needing twenty prepared anecdotes. Six flexible ones, genuinely understood, will cover almost anything you are asked.

The position drill

Draw an ethical prompt and give yourself fifteen seconds to state a clear position in one sentence, before any justification. Then defend it for a minute. Then argue the opposite for a minute.

The most common failure in ethics stations is not being wrong. It is spending six minutes acknowledging that both sides have merit and never landing anywhere. This drill makes landing automatic, and the opposite side round keeps you from becoming rigid.

The empty chair roleplay

Roleplay is the hardest thing to practise alone, and the workaround is imperfect but useful. Put a chair opposite you, decide who is in it and what they are upset about, and talk to them out loud, leaving real gaps where their answers would go. Record it. On playback, ask whether your questions could have been written before the scene started, or whether they responded to something.

You can also record a voice memo playing the other person first, with pauses, then respond to it live. It feels ridiculous. It also builds the habit of listening before advising, which is the behaviour the station is built to test.

Where your prompts come from

Ignore anything advertised as real or leaked questions from last year. It is unverifiable, frequently invented, and universities change their stations anyway. You do not need the actual questions, you need the actual categories.

Write your own instead. Good sources are everywhere once you look:

• Health stories in the news, turned into a question about how you would weigh the competing interests.
• Any conflict you have witnessed at work, converted into a scenario with the names removed.
• The values statements universities publish about the doctors they want, each one turned into a question.
• Rural, remote and equity of access issues, which recur constantly in Australian and New Zealand selection.
• Your own application. Every line of it is a legitimate question, and you should have said each answer out loud at least once.

Practise in the format you will actually face

If your interview runs online, do your solo drills to a camera rather than to a wall, because the skills are not identical. If it runs in person, stand up and walk into the room between prompts. Our comparison of virtual and in person MMIs sets out what changes between the two.

Once a week, do a full dress rehearsal: real clothes from our guide to what to wear to a medical interview, real time of day, four or five stations back to back with short gaps. Stamina is a separate skill from answering, and running one station at a time never reveals that you fall apart by the fifth.

The ceiling, and how to get past it

Two things solo work genuinely cannot give you. The first is a response you did not expect, which is what a real actor or interviewer provides. The second is an outside judgement, because you will always mark yourself generously on the things you did not know were weaknesses.

That is the gap MasterMed's live AI interviewer is built for: timed MMI stations that talk back, marked against a rubric, so you are not the only judge of your own performance. The first speaking station is free on the trial, there is no card, and the trial does not convert by itself.

Until then, the rule is simple. Out loud, on a clock, recorded, on prompts you did not choose. Twenty minutes a day of that beats three hours of reading model answers, and it is available to you tonight, in your bedroom, with nobody else involved.